CV_Jupyter_Notebook_Final.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"https://radar.community.uaf.edu/wp-content/uploads/sites/667/2021/03/HydroSARbanner.jpg\" width=\"100%\" />"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Coefficent of Variation Jupyter Notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Python code to implement coefficient of variation (CV) for crop/non-crop classification using a Receiver Operating Characteristic (ROC) curve for a time-series of SAR images. The notebook statistically calculates the CV for a stack of time-series imagery. The CV output is then used to generate a ROC curve by using the USDA Cropland Data Layer (CDL) as ground truth. Pixels classified by the CDL as \"Water\" are masked and not used in classification because, water has a high variation measurement not comparable to the CV values of other non-cropland land covers and is often missclassified because of this. The statistic Youden's Index is calculated to detemine the ideal threshold on the curve to use for best classification results. The accuracy of the classification compared to the CDL as ground truth are calculated for the CV crop/non-crop classification. The classified image is exported as a Geotiff. A CSV file is exported containing accuracy statistics for the classification.\n",
    "\n",
    "Coefficient of Variation is calculated by: Standard Deviation/Mean\n",
    "\n",
    "Created by: Shannon Rose, Microwave Remote Sensing Labratory (MIRSL) Unuversity of Massachusetts- Amherst\n",
    "\n",
    "Datasets needed:\n",
    "\n",
    "    Timeseries of SAR imagery\n",
    "    \n",
    "    CDL available at https://nassgeodata.gmu.edu/CropScape/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Flow Model for Notebook"
   ]
  },
  {
   "attachments": {
    "CV_ROC_Data_Flow.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![CV_ROC_Data_Flow.png](attachment:CV_ROC_Data_Flow.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Import Python Modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import glob\n",
    "import matplotlib.pyplot as plt\n",
    "from osgeo import gdal, osr\n",
    "import os\n",
    "import pandas as pd\n",
    "from pandas import DataFrame\n",
    "from IPython.display import Image\n",
    "import sklearn\n",
    "from sklearn import metrics\n",
    "from matplotlib.pyplot import figure\n",
    "%matplotlib inline\n",
    "import matplotlib.patches as mpatches"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Set the Input Parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Note: \n",
    "    1. For use in this notebook all raster datasets must be set to the same grid with pixels aligned\n",
    "    2. The only data that can be in the set \"SAR_image_folder\" is the timeseries SAR imagery"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "SAR_image_folder = r'C:\\Users\\sjrose\\Box\\Shannon_Final_Deliverables\\Final_CV_notebook_and_test_data\\Sample_Sentinel1_Time_Series'\n",
    "CDL_image_folder = r'C:\\Users\\sjrose\\Box\\Shannon_Final_Deliverables\\Final_CV_notebook_and_test_data\\Sample_CDL'\n",
    "CDL = \"CDL_Kansas_clip_prep.tif\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Set the Output Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "Output_folder = r'C:\\Users\\sjrose\\Box\\Shannon_Final_Deliverables\\Final_CV_notebook_and_test_data\\Notebook_Outputs'\n",
    "CV_classification_binary = \"CV_classification_final.tif\"\n",
    "CV_classification_filename_with_accuracy = \"CV_classification_with_accuracy.tif\"\n",
    "Accuracy_results = 'accuracy_statistics.csv'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# CDL Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Change working directory to bring in CDL\n",
    "os.chdir(CDL_image_folder)\n",
    "\n",
    "# Open CDL\n",
    "rasterstack = gdal.Open(CDL).ReadAsArray()\n",
    "\n",
    "# Get dimensions of CDL\n",
    "first_raster_CDL = gdal.Open(CDL)\n",
    "rows1 = first_raster_CDL.RasterYSize\n",
    "cols1 = first_raster_CDL.RasterXSize\n",
    "\n",
    "CDL_geotransform = first_raster_CDL.GetGeoTransform()\n",
    "CDL_xres = CDL_geotransform[1]\n",
    "CDL_yres = -CDL_geotransform[5]\n",
    "CDL_xmin = CDL_geotransform[0]\n",
    "CDL_ymax = CDL_geotransform[3]\n",
    "CDL_spatialref = first_raster_CDL.GetProjectionRef()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Creating a mask to not include pixels in the CDL classified as \"Water\".\n",
    "Water has a high variation measurement not comparable to the CV values of other non-cropland land covers and is often missclassified because of this.\n",
    "\n",
    "The numeric classification value of \"Water\" in the CDL is 111.\n",
    "The mask sets water (111 values) to 111 and everything else to 0.\n",
    "\n",
    "Note: Additional masking can be completed outside of this notebook by classifying these areas in the CDL as 111 for use in this         notebook.\n",
    "   Some pervious factors that were masked during classification to eliminate errors inherently present in the CDL:\n",
    "       1. Pixels that are not surrounded by their same land cover type \n",
    "       2. Classified land areas smaller than two hectares in size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "water_number = 111\n",
    "watermask1 = rasterstack == water_number\n",
    "watermask= np.multiply(rasterstack,watermask1)\n",
    "\n",
    "#getting the number of pixels masked and not masked \n",
    "cdl111 = np.count_nonzero(watermask == 111)\n",
    "print(\"Number of masked Pixels:\", cdl111)\n",
    "cdl0 = np.count_nonzero(watermask == 0)\n",
    "print(\"Number of pixels for classification:\", cdl0)\n",
    "plt.imshow(rasterstack)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Creating a binary crop/non-crop classification from the CDL classified classes.\n",
    "\n",
    "Crop classified to 1\n",
    "\n",
    "Non-crop classified to 0\n",
    "\n",
    "This can be customized by study area and what land use types are present there. \n",
    "To make a land class classified as non-crop simply place a \"#\" infront of the line of code for that land cover"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "CDL_crop = np.copy(rasterstack)\n",
    "unique = np.unique(rasterstack)\n",
    "uniquecount = len(unique)\n",
    "\n",
    "#setting all crop classification values equal to 1 \n",
    "CDL_crop[np.where(rasterstack== 1)]= 1    #Corn\n",
    "CDL_crop[np.where(rasterstack== 2)]= 1    #Cotton\n",
    "CDL_crop[np.where(rasterstack== 3)]= 1    #Rice\n",
    "CDL_crop[np.where(rasterstack== 4)]= 1    #Sorghum\n",
    "CDL_crop[np.where(rasterstack== 5)]= 1    #Soybeans\n",
    "CDL_crop[np.where(rasterstack== 6)]= 1    #Sunflower\n",
    "CDL_crop[np.where(rasterstack== 10)]= 1    #Peanuts\n",
    "CDL_crop[np.where(rasterstack== 11)]= 1    #Tobacco\n",
    "CDL_crop[np.where(rasterstack== 12)]= 1    #Sweet Corn\n",
    "CDL_crop[np.where(rasterstack== 13)]= 1    #Pop or Orn Corn\n",
    "CDL_crop[np.where(rasterstack== 14)]= 1    #Mint\n",
    "CDL_crop[np.where(rasterstack== 21)]= 1    #Barley\n",
    "CDL_crop[np.where(rasterstack== 22)]= 1    #Durum Wheat\n",
    "CDL_crop[np.where(rasterstack== 23)]= 1    #Spring Wheat\n",
    "CDL_crop[np.where(rasterstack== 24)]= 1    #Winter Wheat\n",
    "CDL_crop[np.where(rasterstack== 25)]= 1    #Other small grains\n",
    "CDL_crop[np.where(rasterstack== 26)]= 1    #Dbl crop WinWht/Soybeans\n",
    "CDL_crop[np.where(rasterstack== 27)]= 1    #Rye\n",
    "CDL_crop[np.where(rasterstack== 28)]= 1    #Oats\n",
    "CDL_crop[np.where(rasterstack== 29)]= 1    #Millet\n",
    "CDL_crop[np.where(rasterstack== 30)]= 1    #Speltz\n",
    "CDL_crop[np.where(rasterstack== 31)]= 1    #Canola\n",
    "CDL_crop[np.where(rasterstack== 32)]= 1    #Flaxseed\n",
    "CDL_crop[np.where(rasterstack== 33)]= 1    #Safflower\n",
    "CDL_crop[np.where(rasterstack== 34)]= 1    #Rape Seed\n",
    "CDL_crop[np.where(rasterstack== 35)]= 1    #Mustard\n",
    "CDL_crop[np.where(rasterstack== 36)]= 1    #Alfalfa\n",
    "#CDL_crop[np.where(rasterstack== 37)]= 1    #Other Hay/Non Alfalfa---- could be omitted\n",
    "CDL_crop[np.where(rasterstack== 38)]= 1    #Camelina\n",
    "CDL_crop[np.where(rasterstack== 39)]= 1    #Buckwheat\n",
    "CDL_crop[np.where(rasterstack== 41)]= 1    #Sugarbeets\n",
    "CDL_crop[np.where(rasterstack== 42)]= 1    #Dry Beans\n",
    "CDL_crop[np.where(rasterstack== 43)]= 1    #Potatoes\n",
    "CDL_crop[np.where(rasterstack== 44)]= 1    #Other Crops\n",
    "CDL_crop[np.where(rasterstack== 45)]= 1    #Sugarcane\n",
    "CDL_crop[np.where(rasterstack== 46)]= 1    #Sweet Potatoes\n",
    "CDL_crop[np.where(rasterstack== 47)]= 1    #Misc Vegs & Fruits\n",
    "CDL_crop[np.where(rasterstack== 48)]= 1    #Watermelons\n",
    "CDL_crop[np.where(rasterstack== 49)]= 1    #Onions\n",
    "CDL_crop[np.where(rasterstack== 50)]= 1    #Cucumbers\n",
    "CDL_crop[np.where(rasterstack== 51)]= 1    #Chick Peas\n",
    "CDL_crop[np.where(rasterstack== 52)]= 1    #Lentils\n",
    "CDL_crop[np.where(rasterstack== 53)]= 1    #Peas\n",
    "CDL_crop[np.where(rasterstack== 54)]= 1    #Tomatoes\n",
    "CDL_crop[np.where(rasterstack== 55)]= 1    #Cranberries\n",
    "CDL_crop[np.where(rasterstack== 56)]= 1    #Hops\n",
    "CDL_crop[np.where(rasterstack== 57)]= 1    #Herbs\n",
    "CDL_crop[np.where(rasterstack== 58)]= 1    #Clover/Wildflowers\n",
    "#CDL_crop[np.where(rasterstack== 59)]= 1    #Sod/Grass Seed---- could be omitted\n",
    "#CDL_crop[np.where(rasterstack== 60)]= 1    #Switchgrass---- could be omitted\n",
    "CDL_crop[np.where(rasterstack== 61)]= 1    #Fallow/Idle Cropland\n",
    "CDL_crop[np.where(rasterstack== 66)]= 1    #Cherries\n",
    "CDL_crop[np.where(rasterstack== 67)]= 1    #Peaches\n",
    "CDL_crop[np.where(rasterstack== 68)]= 1    #Apples\n",
    "CDL_crop[np.where(rasterstack== 69)]= 1    #Grapes\n",
    "CDL_crop[np.where(rasterstack== 70)]= 1    #Christmas Trees\n",
    "CDL_crop[np.where(rasterstack== 71)]= 1    #Other Tree Crops    \n",
    "CDL_crop[np.where(rasterstack== 72)]= 1    #Citrus\n",
    "CDL_crop[np.where(rasterstack== 74)]= 1    #Pecans\n",
    "CDL_crop[np.where(rasterstack== 75)]= 1    #Almonds\n",
    "CDL_crop[np.where(rasterstack== 76)]= 1    #Walnuts\n",
    "CDL_crop[np.where(rasterstack== 77)]= 1    #Pears\n",
    "#CDL_crop[np.where(rasterstack== 176)]= 1   #Grassland/Pasture\n",
    "#CDL_crop[np.where(rasterstack== 121)]= 1    #Developed/Open Space\n",
    "CDL_crop[np.where(rasterstack== 204)]= 1    #Pistachios\n",
    "CDL_crop[np.where(rasterstack== 205)]= 1    #Triticale\n",
    "CDL_crop[np.where(rasterstack== 206)]= 1    #Carrots\n",
    "CDL_crop[np.where(rasterstack== 207)]= 1    #Asparagus\n",
    "CDL_crop[np.where(rasterstack== 208)]= 1    #Garlic\n",
    "CDL_crop[np.where(rasterstack== 209)]= 1    #Cantaloupes\n",
    "CDL_crop[np.where(rasterstack== 210)]= 1    #Prunes\n",
    "CDL_crop[np.where(rasterstack== 211)]= 1    #Olives\n",
    "CDL_crop[np.where(rasterstack== 212)]= 1    #Oranges\n",
    "CDL_crop[np.where(rasterstack== 213)]= 1    #Honeydew Melons\n",
    "CDL_crop[np.where(rasterstack== 214)]= 1    #Broccoli\n",
    "CDL_crop[np.where(rasterstack== 216)]= 1    #Peppers\n",
    "CDL_crop[np.where(rasterstack== 217)]= 1    #Pomegranates\n",
    "CDL_crop[np.where(rasterstack== 218)]= 1    #Nectarines\n",
    "CDL_crop[np.where(rasterstack== 219)]= 1    #Greens\n",
    "CDL_crop[np.where(rasterstack== 220)]= 1    #Plums\n",
    "CDL_crop[np.where(rasterstack== 221)]= 1    #Strawberries\n",
    "CDL_crop[np.where(rasterstack== 222)]= 1    #Squash\n",
    "CDL_crop[np.where(rasterstack== 223)]= 1    #Apricots\n",
    "CDL_crop[np.where(rasterstack== 224)]= 1    #Vetch\n",
    "CDL_crop[np.where(rasterstack== 225)]= 1    #Dbl Crop WinWht/Corn\n",
    "CDL_crop[np.where(rasterstack== 226)]= 1    #Dbl Crop Oats/Corn\n",
    "CDL_crop[np.where(rasterstack== 227)]= 1    #Lettuce\n",
    "CDL_crop[np.where(rasterstack== 229)]= 1    #Pumpkins\n",
    "CDL_crop[np.where(rasterstack== 230)]= 1    #Dbl Crop Lettuce/Drum Wht\n",
    "CDL_crop[np.where(rasterstack== 231)]= 1    #Dbl Crop Lettuce/Cantaloupe\n",
    "CDL_crop[np.where(rasterstack== 232)]= 1    #Dbl Crop Lettuce/Cotton\n",
    "CDL_crop[np.where(rasterstack== 233)]= 1    #Dbl Crop Lettuce/Barley\n",
    "CDL_crop[np.where(rasterstack== 234)]= 1    #Dbl Crop Durum Wht/Sorghum\n",
    "CDL_crop[np.where(rasterstack== 235)]= 1    #Dbl Crop Barley/Sorghum\n",
    "CDL_crop[np.where(rasterstack== 236)]= 1    #Dbl Crop WinWht/Sorghum\n",
    "CDL_crop[np.where(rasterstack== 237)]= 1    #Dbl Crop Barley/Corn\n",
    "CDL_crop[np.where(rasterstack== 238)]= 1    #Dbl Crop WinWht/Cotton\n",
    "CDL_crop[np.where(rasterstack== 239)]= 1    #Dbl Crop Soybeans/Cotton\n",
    "CDL_crop[np.where(rasterstack== 240)]= 1    #Dbl Crop Soybeans/Oats\n",
    "CDL_crop[np.where(rasterstack== 241)]= 1    #Dbl Crop Corn/Soybeans\n",
    "CDL_crop[np.where(rasterstack== 242)]= 1    #Blueberries\n",
    "CDL_crop[np.where(rasterstack== 243)]= 1    #Cabbage\n",
    "CDL_crop[np.where(rasterstack== 244)]= 1    #Cauliflower\n",
    "CDL_crop[np.where(rasterstack== 245)]= 1    #Celery\n",
    "CDL_crop[np.where(rasterstack== 246)]= 1    #Radishes\n",
    "CDL_crop[np.where(rasterstack== 247)]= 1    #Turnips\n",
    "CDL_crop[np.where(rasterstack== 248)]= 1    #Eggplants\n",
    "CDL_crop[np.where(rasterstack== 249)]= 1    #Gourds\n",
    "CDL_crop[np.where(rasterstack== 250)]= 1    #Cranberries\n",
    "CDL_crop[np.where(rasterstack== 254)]= 1    #Dbl Crop Barley/Soybeans\n",
    "\n",
    "\n",
    "#setting all non-crop values to 0 based on assumming that everything not previously classified as crop is non-crop\n",
    "CDL_binary = np.copy(CDL_crop)\n",
    "CDL_binary[np.where(CDL_crop!= 1)]= 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Finding the breakdown of crop/non-crop pixels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#getting the number of pixels classified as crop and non-crop including the removal of masked pixels\n",
    "cropcount = np.count_nonzero(CDL_binary == 1)\n",
    "print(\"Number of crop pixels:\", cropcount)\n",
    "\n",
    "noncount = np.count_nonzero(CDL_binary == 0)- cdl111\n",
    "print(\"Number of non-crop pixels:\",noncount)\n",
    "\n",
    "#finding the percent of pixels classified as crop and non-crop including the removal of masked pixels\n",
    "percent_crop_nonmasked_calc = round((cropcount / cdl0)*100, 2)\n",
    "print (\"% Crop: \", percent_crop_nonmasked_calc)\n",
    "\n",
    "percent_noncrop_nonmasked_calc = round((noncount / cdl0)*100, 2)\n",
    "print (\"% Non-crop: \", percent_noncrop_nonmasked_calc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Bringing in SAR Images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Create a timeseries stack of the SAR imagery"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Setting the working directory \n",
    "os.chdir(SAR_image_folder)\n",
    "\n",
    "\n",
    "# Selecting the images for the time series-- all \"\".tif\" images in set working directory\n",
    "SAR_images = glob.glob(\"*tif\")\n",
    "\n",
    "\n",
    "# Read raster files and make them into a 3D numpy array\n",
    "arrs = []\n",
    "for image in SAR_images:\n",
    "    arrs.append(gdal.Open(image).ReadAsArray())\n",
    "a = np.array(arrs, dtype=np.float)\n",
    "\n",
    "\n",
    "# Creating a variable for the number of images (one for each date)\n",
    "num_dates = len(arrs)\n",
    "print (\"Number of dates:\", num_dates)\n",
    "\n",
    "\n",
    "# Get dimensions of the first SAR image--- should be the same for every image in the stack and the CDL\n",
    "first_raster = gdal.Open(SAR_images[1])\n",
    "rows = first_raster.RasterYSize\n",
    "cols = first_raster.RasterXSize\n",
    "geotransform = first_raster.GetGeoTransform()\n",
    "xres = geotransform[1]\n",
    "yres = -geotransform[5]\n",
    "xmin = geotransform[0]\n",
    "ymax = geotransform[3]\n",
    "spatialref = first_raster.GetProjectionRef()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Calculating the CV on the Timeseries Stack"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "CV = Standard Deviation / Mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Calculate the mean for the time stack of images\n",
    "mean = np.mean(a, axis = 0)\n",
    "\n",
    "\n",
    "# Calculate the standard deviation for the time stack of images\n",
    "std = np.std(a, axis = 0)\n",
    "\n",
    "\n",
    "# Calculate the coefficient of variation for the time stack of images \n",
    "CV = std/mean"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Generation of the ROC Curve"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "To create the curve 100 different sensitivity points between 0 and 100 were tested. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Reclassify CV array to distingush crop from non-crop\n",
    "\n",
    "#0.0\n",
    "CV_reclass_0 = np.copy(CV)\n",
    "CV_reclass_0[np.where(CV_reclass_0>= 0.0)] = 1\n",
    "CV_reclass_0[np.where(CV_reclass_0< 0.0)] = 0\n",
    "\n",
    "CV_reclass_01 = np.copy(CV)\n",
    "CV_reclass_01[np.where(CV_reclass_01>= 0.01)] = 1\n",
    "CV_reclass_01[np.where(CV_reclass_01< 0.01)] = 0\n",
    "\n",
    "CV_reclass_02 = np.copy(CV)\n",
    "CV_reclass_02[np.where(CV_reclass_02>= 0.02)] = 1\n",
    "CV_reclass_02[np.where(CV_reclass_02< 0.02)] = 0\n",
    "\n",
    "CV_reclass_03 = np.copy(CV)\n",
    "CV_reclass_03[np.where(CV_reclass_03>= 0.03)] = 1\n",
    "CV_reclass_03[np.where(CV_reclass_03< 0.03)] = 0\n",
    "\n",
    "CV_reclass_04 = np.copy(CV)\n",
    "CV_reclass_04[np.where(CV_reclass_04>= 0.04)] = 1\n",
    "CV_reclass_04[np.where(CV_reclass_04< 0.04)] = 0\n",
    "\n",
    "CV_reclass_05 = np.copy(CV)\n",
    "CV_reclass_05[np.where(CV_reclass_05>= 0.05)] = 1\n",
    "CV_reclass_05[np.where(CV_reclass_05< 0.05)] = 0\n",
    "\n",
    "CV_reclass_06 = np.copy(CV)\n",
    "CV_reclass_06[np.where(CV_reclass_06>= 0.06)] = 1\n",
    "CV_reclass_06[np.where(CV_reclass_06< 0.06)] = 0\n",
    "\n",
    "CV_reclass_07 = np.copy(CV)\n",
    "CV_reclass_07[np.where(CV_reclass_07>= 0.07)] = 1\n",
    "CV_reclass_07[np.where(CV_reclass_07< 0.07)] = 0\n",
    "\n",
    "CV_reclass_08 = np.copy(CV)\n",
    "CV_reclass_08[np.where(CV_reclass_08>= 0.08)] = 1\n",
    "CV_reclass_08[np.where(CV_reclass_08< 0.08)] = 0\n",
    "\n",
    "CV_reclass_09 = np.copy(CV)\n",
    "CV_reclass_09[np.where(CV_reclass_09>= 0.09)] = 1\n",
    "CV_reclass_09[np.where(CV_reclass_09< 0.09)] = 0\n",
    "\n",
    "#0.1\n",
    "CV_reclass_1 = np.copy(CV)\n",
    "CV_reclass_1[np.where(CV_reclass_1>= 0.1)] = 1\n",
    "CV_reclass_1[np.where(CV_reclass_1< 0.1)] = 0\n",
    "\n",
    "CV_reclass_11 = np.copy(CV)\n",
    "CV_reclass_11[np.where(CV_reclass_11>= 0.11)] = 1\n",
    "CV_reclass_11[np.where(CV_reclass_11< 0.11)] = 0\n",
    "\n",
    "CV_reclass_12 = np.copy(CV)\n",
    "CV_reclass_12[np.where(CV_reclass_12>= 0.12)] = 1\n",
    "CV_reclass_12[np.where(CV_reclass_12< 0.12)] = 0\n",
    "\n",
    "CV_reclass_13 = np.copy(CV)\n",
    "CV_reclass_13[np.where(CV_reclass_13>= 0.13)] = 1\n",
    "CV_reclass_13[np.where(CV_reclass_13< 0.13)] = 0\n",
    "\n",
    "CV_reclass_14 = np.copy(CV)\n",
    "CV_reclass_14[np.where(CV_reclass_14>= 0.14)] = 1\n",
    "CV_reclass_14[np.where(CV_reclass_14< 0.14)] = 0\n",
    "\n",
    "CV_reclass_15 = np.copy(CV)\n",
    "CV_reclass_15[np.where(CV_reclass_15>= 0.15)] = 1\n",
    "CV_reclass_15[np.where(CV_reclass_15< 0.15)] = 0\n",
    "\n",
    "CV_reclass_16 = np.copy(CV)\n",
    "CV_reclass_16[np.where(CV_reclass_16>= 0.16)] = 1\n",
    "CV_reclass_16[np.where(CV_reclass_16< 0.16)] = 0\n",
    "\n",
    "CV_reclass_17 = np.copy(CV)\n",
    "CV_reclass_17[np.where(CV_reclass_17>= 0.17)] = 1\n",
    "CV_reclass_17[np.where(CV_reclass_17< 0.17)] = 0\n",
    "\n",
    "CV_reclass_18 = np.copy(CV)\n",
    "CV_reclass_18[np.where(CV_reclass_18>= 0.18)] = 1\n",
    "CV_reclass_18[np.where(CV_reclass_18< 0.18)] = 0\n",
    "\n",
    "CV_reclass_19 = np.copy(CV)\n",
    "CV_reclass_19[np.where(CV_reclass_19>= 0.19)] = 1\n",
    "CV_reclass_19[np.where(CV_reclass_19< 0.19)] = 0\n",
    "\n",
    "#0.2\n",
    "CV_reclass_2 = np.copy(CV)\n",
    "CV_reclass_2[np.where(CV_reclass_2>= 0.2)] = 1\n",
    "CV_reclass_2[np.where(CV_reclass_2< 0.2)] = 0\n",
    "\n",
    "CV_reclass_21 = np.copy(CV)\n",
    "CV_reclass_21[np.where(CV_reclass_21>= 0.21)] = 1\n",
    "CV_reclass_21[np.where(CV_reclass_21< 0.21)] = 0\n",
    "\n",
    "CV_reclass_22 = np.copy(CV)\n",
    "CV_reclass_22[np.where(CV_reclass_22>= 0.22)] = 1\n",
    "CV_reclass_22[np.where(CV_reclass_22< 0.22)] = 0\n",
    "\n",
    "CV_reclass_23 = np.copy(CV)\n",
    "CV_reclass_23[np.where(CV_reclass_23>= 0.23)] = 1\n",
    "CV_reclass_23[np.where(CV_reclass_23< 0.23)] = 0\n",
    "\n",
    "CV_reclass_24 = np.copy(CV)\n",
    "CV_reclass_24[np.where(CV_reclass_24>= 0.24)] = 1\n",
    "CV_reclass_24[np.where(CV_reclass_24< 0.24)] = 0\n",
    "\n",
    "CV_reclass_25 = np.copy(CV)\n",
    "CV_reclass_25[np.where(CV_reclass_25>= 0.25)] = 1\n",
    "CV_reclass_25[np.where(CV_reclass_25< 0.25)] = 0\n",
    "\n",
    "CV_reclass_26 = np.copy(CV)\n",
    "CV_reclass_26[np.where(CV_reclass_26>= 0.26)] = 1\n",
    "CV_reclass_26[np.where(CV_reclass_26< 0.26)] = 0\n",
    "\n",
    "CV_reclass_27 = np.copy(CV)\n",
    "CV_reclass_27[np.where(CV_reclass_27>= 0.27)] = 1\n",
    "CV_reclass_27[np.where(CV_reclass_27< 0.27)] = 0\n",
    "\n",
    "CV_reclass_28 = np.copy(CV)\n",
    "CV_reclass_28[np.where(CV_reclass_28>= 0.28)] = 1\n",
    "CV_reclass_28[np.where(CV_reclass_28< 0.28)] = 0\n",
    "\n",
    "CV_reclass_29 = np.copy(CV)\n",
    "CV_reclass_29[np.where(CV_reclass_29>= 0.29)] = 1\n",
    "CV_reclass_29[np.where(CV_reclass_29< 0.29)] = 0\n",
    "\n",
    "#0.3\n",
    "CV_reclass_3 = np.copy(CV)\n",
    "CV_reclass_3[np.where(CV_reclass_3>= 0.3)] = 1\n",
    "CV_reclass_3[np.where(CV_reclass_3< 0.3)] = 0\n",
    "\n",
    "CV_reclass_31 = np.copy(CV)\n",
    "CV_reclass_31[np.where(CV_reclass_31>= 0.31)] = 1\n",
    "CV_reclass_31[np.where(CV_reclass_31< 0.31)] = 0\n",
    "\n",
    "CV_reclass_32 = np.copy(CV)\n",
    "CV_reclass_32[np.where(CV_reclass_32>= 0.32)] = 1\n",
    "CV_reclass_32[np.where(CV_reclass_32< 0.32)] = 0\n",
    "\n",
    "CV_reclass_33 = np.copy(CV)\n",
    "CV_reclass_33[np.where(CV_reclass_33>= 0.33)] = 1\n",
    "CV_reclass_33[np.where(CV_reclass_33< 0.33)] = 0\n",
    "\n",
    "CV_reclass_34 = np.copy(CV)\n",
    "CV_reclass_34[np.where(CV_reclass_34>= 0.34)] = 1\n",
    "CV_reclass_34[np.where(CV_reclass_34< 0.34)] = 0\n",
    "\n",
    "CV_reclass_35 = np.copy(CV)\n",
    "CV_reclass_35[np.where(CV_reclass_35>= 0.35)] = 1\n",
    "CV_reclass_35[np.where(CV_reclass_35< 0.35)] = 0\n",
    "\n",
    "CV_reclass_36 = np.copy(CV)\n",
    "CV_reclass_36[np.where(CV_reclass_36>= 0.36)] = 1\n",
    "CV_reclass_36[np.where(CV_reclass_36< 0.36)] = 0\n",
    "\n",
    "CV_reclass_37 = np.copy(CV)\n",
    "CV_reclass_37[np.where(CV_reclass_37>= 0.37)] = 1\n",
    "CV_reclass_37[np.where(CV_reclass_37< 0.37)] = 0\n",
    "\n",
    "CV_reclass_38 = np.copy(CV)\n",
    "CV_reclass_38[np.where(CV_reclass_38>= 0.38)] = 1\n",
    "CV_reclass_38[np.where(CV_reclass_38< 0.38)] = 0\n",
    "\n",
    "CV_reclass_39 = np.copy(CV)\n",
    "CV_reclass_39[np.where(CV_reclass_39>= 0.39)] = 1\n",
    "CV_reclass_39[np.where(CV_reclass_39< 0.39)] = 0\n",
    "\n",
    "#0.4\n",
    "CV_reclass_4 = np.copy(CV)\n",
    "CV_reclass_4[np.where(CV_reclass_4>= 0.4)] = 1\n",
    "CV_reclass_4[np.where(CV_reclass_4< 0.4)] = 0\n",
    "\n",
    "CV_reclass_41 = np.copy(CV)\n",
    "CV_reclass_41[np.where(CV_reclass_41>= 0.41)] = 1\n",
    "CV_reclass_41[np.where(CV_reclass_41< 0.41)] = 0\n",
    "\n",
    "CV_reclass_42 = np.copy(CV)\n",
    "CV_reclass_42[np.where(CV_reclass_42>= 0.42)] = 1\n",
    "CV_reclass_42[np.where(CV_reclass_42< 0.42)] = 0\n",
    "\n",
    "CV_reclass_43 = np.copy(CV)\n",
    "CV_reclass_43[np.where(CV_reclass_43>= 0.43)] = 1\n",
    "CV_reclass_43[np.where(CV_reclass_43< 0.43)] = 0\n",
    "\n",
    "CV_reclass_44 = np.copy(CV)\n",
    "CV_reclass_44[np.where(CV_reclass_44>= 0.44)] = 1\n",
    "CV_reclass_44[np.where(CV_reclass_44< 0.44)] = 0\n",
    "\n",
    "CV_reclass_45 = np.copy(CV)\n",
    "CV_reclass_45[np.where(CV_reclass_45>= 0.45)] = 1\n",
    "CV_reclass_45[np.where(CV_reclass_45< 0.45)] = 0\n",
    "\n",
    "CV_reclass_46 = np.copy(CV)\n",
    "CV_reclass_46[np.where(CV_reclass_46>= 0.46)] = 1\n",
    "CV_reclass_46[np.where(CV_reclass_46< 0.46)] = 0\n",
    "\n",
    "CV_reclass_47 = np.copy(CV)\n",
    "CV_reclass_47[np.where(CV_reclass_47>= 0.47)] = 1\n",
    "CV_reclass_47[np.where(CV_reclass_47< 0.47)] = 0\n",
    "\n",
    "CV_reclass_48 = np.copy(CV)\n",
    "CV_reclass_48[np.where(CV_reclass_48>= 0.48)] = 1\n",
    "CV_reclass_48[np.where(CV_reclass_48< 0.48)] = 0\n",
    "\n",
    "CV_reclass_49 = np.copy(CV)\n",
    "CV_reclass_49[np.where(CV_reclass_49>= 0.49)] = 1\n",
    "CV_reclass_49[np.where(CV_reclass_49< 0.49)] = 0\n",
    "\n",
    "#0.5\n",
    "CV_reclass_5 = np.copy(CV)\n",
    "CV_reclass_5[np.where(CV_reclass_5>= 0.5)] = 1\n",
    "CV_reclass_5[np.where(CV_reclass_5< 0.5)] = 0\n",
    "\n",
    "CV_reclass_51 = np.copy(CV)\n",
    "CV_reclass_51[np.where(CV_reclass_51>= 0.51)] = 1\n",
    "CV_reclass_51[np.where(CV_reclass_51< 0.51)] = 0\n",
    "\n",
    "CV_reclass_52 = np.copy(CV)\n",
    "CV_reclass_52[np.where(CV_reclass_52>= 0.52)] = 1\n",
    "CV_reclass_52[np.where(CV_reclass_52< 0.52)] = 0\n",
    "\n",
    "CV_reclass_53 = np.copy(CV)\n",
    "CV_reclass_53[np.where(CV_reclass_53>= 0.53)] = 1\n",
    "CV_reclass_53[np.where(CV_reclass_53< 0.53)] = 0\n",
    "\n",
    "CV_reclass_54 = np.copy(CV)\n",
    "CV_reclass_54[np.where(CV_reclass_54>= 0.54)] = 1\n",
    "CV_reclass_54[np.where(CV_reclass_54< 0.54)] = 0\n",
    "\n",
    "CV_reclass_55 = np.copy(CV)\n",
    "CV_reclass_55[np.where(CV_reclass_55>= 0.55)] = 1\n",
    "CV_reclass_55[np.where(CV_reclass_55< 0.55)] = 0\n",
    "\n",
    "CV_reclass_56 = np.copy(CV)\n",
    "CV_reclass_56[np.where(CV_reclass_56>= 0.56)] = 1\n",
    "CV_reclass_56[np.where(CV_reclass_56< 0.56)] = 0\n",
    "\n",
    "CV_reclass_57 = np.copy(CV)\n",
    "CV_reclass_57[np.where(CV_reclass_57>= 0.57)] = 1\n",
    "CV_reclass_57[np.where(CV_reclass_57< 0.57)] = 0\n",
    "\n",
    "CV_reclass_58 = np.copy(CV)\n",
    "CV_reclass_58[np.where(CV_reclass_58>= 0.58)] = 1\n",
    "CV_reclass_58[np.where(CV_reclass_58< 0.58)] = 0\n",
    "\n",
    "CV_reclass_59 = np.copy(CV)\n",
    "CV_reclass_59[np.where(CV_reclass_59>= 0.59)] = 1\n",
    "CV_reclass_59[np.where(CV_reclass_59< 0.59)] = 0\n",
    "\n",
    "#0.6\n",
    "CV_reclass_6 = np.copy(CV)\n",
    "CV_reclass_6[np.where(CV_reclass_6>= 0.6)] = 1\n",
    "CV_reclass_6[np.where(CV_reclass_6< 0.6)] = 0\n",
    "\n",
    "CV_reclass_61 = np.copy(CV)\n",
    "CV_reclass_61[np.where(CV_reclass_61>= 0.61)] = 1\n",
    "CV_reclass_61[np.where(CV_reclass_61< 0.61)] = 0\n",
    "\n",
    "CV_reclass_62 = np.copy(CV)\n",
    "CV_reclass_62[np.where(CV_reclass_62>= 0.62)] = 1\n",
    "CV_reclass_62[np.where(CV_reclass_62< 0.62)] = 0\n",
    "\n",
    "CV_reclass_63 = np.copy(CV)\n",
    "CV_reclass_63[np.where(CV_reclass_63>= 0.63)] = 1\n",
    "CV_reclass_63[np.where(CV_reclass_63< 0.63)] = 0\n",
    "\n",
    "CV_reclass_64 = np.copy(CV)\n",
    "CV_reclass_64[np.where(CV_reclass_64>= 0.64)] = 1\n",
    "CV_reclass_64[np.where(CV_reclass_64< 0.64)] = 0\n",
    "\n",
    "CV_reclass_65 = np.copy(CV)\n",
    "CV_reclass_65[np.where(CV_reclass_65>= 0.65)] = 1\n",
    "CV_reclass_65[np.where(CV_reclass_65< 0.65)] = 0\n",
    "\n",
    "CV_reclass_66 = np.copy(CV)\n",
    "CV_reclass_66[np.where(CV_reclass_66>= 0.66)] = 1\n",
    "CV_reclass_66[np.where(CV_reclass_66< 0.66)] = 0\n",
    "\n",
    "CV_reclass_67 = np.copy(CV)\n",
    "CV_reclass_67[np.where(CV_reclass_67>= 0.67)] = 1\n",
    "CV_reclass_67[np.where(CV_reclass_67< 0.67)] = 0\n",
    "\n",
    "CV_reclass_68 = np.copy(CV)\n",
    "CV_reclass_68[np.where(CV_reclass_68>= 0.68)] = 1\n",
    "CV_reclass_68[np.where(CV_reclass_68< 0.68)] = 0\n",
    "\n",
    "CV_reclass_69 = np.copy(CV)\n",
    "CV_reclass_69[np.where(CV_reclass_69>= 0.69)] = 1\n",
    "CV_reclass_69[np.where(CV_reclass_69< 0.69)] = 0\n",
    "\n",
    "#0.7\n",
    "CV_reclass_7 = np.copy(CV)\n",
    "CV_reclass_7[np.where(CV_reclass_7>= 0.7)] = 1\n",
    "CV_reclass_7[np.where(CV_reclass_7< 0.7)] = 0\n",
    "\n",
    "CV_reclass_71 = np.copy(CV)\n",
    "CV_reclass_71[np.where(CV_reclass_71>= 0.71)] = 1\n",
    "CV_reclass_71[np.where(CV_reclass_71< 0.71)] = 0\n",
    "\n",
    "CV_reclass_72 = np.copy(CV)\n",
    "CV_reclass_72[np.where(CV_reclass_72>= 0.72)] = 1\n",
    "CV_reclass_72[np.where(CV_reclass_72< 0.72)] = 0\n",
    "\n",
    "CV_reclass_73 = np.copy(CV)\n",
    "CV_reclass_73[np.where(CV_reclass_73>= 0.73)] = 1\n",
    "CV_reclass_73[np.where(CV_reclass_73< 0.73)] = 0\n",
    "\n",
    "CV_reclass_74 = np.copy(CV)\n",
    "CV_reclass_74[np.where(CV_reclass_74>= 0.74)] = 1\n",
    "CV_reclass_74[np.where(CV_reclass_74< 0.74)] = 0\n",
    "\n",
    "CV_reclass_75 = np.copy(CV)\n",
    "CV_reclass_75[np.where(CV_reclass_75>= 0.75)] = 1\n",
    "CV_reclass_75[np.where(CV_reclass_75< 0.75)] = 0\n",
    "\n",
    "CV_reclass_76 = np.copy(CV)\n",
    "CV_reclass_76[np.where(CV_reclass_76>= 0.76)] = 1\n",
    "CV_reclass_76[np.where(CV_reclass_76< 0.76)] = 0\n",
    "\n",
    "CV_reclass_77 = np.copy(CV)\n",
    "CV_reclass_77[np.where(CV_reclass_77>= 0.77)] = 1\n",
    "CV_reclass_77[np.where(CV_reclass_77< 0.77)] = 0\n",
    "\n",
    "CV_reclass_78 = np.copy(CV)\n",
    "CV_reclass_78[np.where(CV_reclass_78>= 0.78)] = 1\n",
    "CV_reclass_78[np.where(CV_reclass_78< 0.78)] = 0\n",
    "\n",
    "CV_reclass_79 = np.copy(CV)\n",
    "CV_reclass_79[np.where(CV_reclass_79>= 0.79)] = 1\n",
    "CV_reclass_79[np.where(CV_reclass_79< 0.79)] = 0\n",
    "\n",
    "#0.8\n",
    "CV_reclass_8 = np.copy(CV)\n",
    "CV_reclass_8[np.where(CV_reclass_8>= 0.8)] = 1\n",
    "CV_reclass_8[np.where(CV_reclass_8< 0.8)] = 0\n",
    "\n",
    "CV_reclass_81 = np.copy(CV)\n",
    "CV_reclass_81[np.where(CV_reclass_81>= 0.81)] = 1\n",
    "CV_reclass_81[np.where(CV_reclass_81< 0.81)] = 0\n",
    "\n",
    "CV_reclass_82 = np.copy(CV)\n",
    "CV_reclass_82[np.where(CV_reclass_82>= 0.82)] = 1\n",
    "CV_reclass_82[np.where(CV_reclass_82< 0.82)] = 0\n",
    "\n",
    "CV_reclass_83 = np.copy(CV)\n",
    "CV_reclass_83[np.where(CV_reclass_83>= 0.83)] = 1\n",
    "CV_reclass_83[np.where(CV_reclass_83< 0.83)] = 0\n",
    "\n",
    "CV_reclass_84 = np.copy(CV)\n",
    "CV_reclass_84[np.where(CV_reclass_84>= 0.84)] = 1\n",
    "CV_reclass_84[np.where(CV_reclass_84< 0.84)] = 0\n",
    "\n",
    "CV_reclass_85 = np.copy(CV)\n",
    "CV_reclass_85[np.where(CV_reclass_85>= 0.85)] = 1\n",
    "CV_reclass_85[np.where(CV_reclass_85< 0.85)] = 0\n",
    "\n",
    "CV_reclass_86 = np.copy(CV)\n",
    "CV_reclass_86[np.where(CV_reclass_86>= 0.86)] = 1\n",
    "CV_reclass_86[np.where(CV_reclass_86< 0.86)] = 0\n",
    "\n",
    "CV_reclass_87 = np.copy(CV)\n",
    "CV_reclass_87[np.where(CV_reclass_87>= 0.87)] = 1\n",
    "CV_reclass_87[np.where(CV_reclass_87< 0.87)] = 0\n",
    "\n",
    "CV_reclass_88 = np.copy(CV)\n",
    "CV_reclass_88[np.where(CV_reclass_88>= 0.88)] = 1\n",
    "CV_reclass_88[np.where(CV_reclass_88< 0.88)] = 0\n",
    "\n",
    "CV_reclass_89 = np.copy(CV)\n",
    "CV_reclass_89[np.where(CV_reclass_89>= 0.89)] = 1\n",
    "CV_reclass_89[np.where(CV_reclass_89< 0.89)] = 0\n",
    "\n",
    "#0.9\n",
    "CV_reclass_9 = np.copy(CV)\n",
    "CV_reclass_9[np.where(CV_reclass_9>= 0.9)] = 1\n",
    "CV_reclass_9[np.where(CV_reclass_9< 0.9)] = 0\n",
    "\n",
    "CV_reclass_91 = np.copy(CV)\n",
    "CV_reclass_91[np.where(CV_reclass_91>= 0.91)] = 1\n",
    "CV_reclass_91[np.where(CV_reclass_91< 0.91)] = 0\n",
    "\n",
    "CV_reclass_92 = np.copy(CV)\n",
    "CV_reclass_92[np.where(CV_reclass_92>= 0.92)] = 1\n",
    "CV_reclass_92[np.where(CV_reclass_92< 0.92)] = 0\n",
    "\n",
    "CV_reclass_93 = np.copy(CV)\n",
    "CV_reclass_93[np.where(CV_reclass_93>= 0.93)] = 1\n",
    "CV_reclass_93[np.where(CV_reclass_93< 0.93)] = 0\n",
    "\n",
    "CV_reclass_94 = np.copy(CV)\n",
    "CV_reclass_94[np.where(CV_reclass_94>= 0.94)] = 1\n",
    "CV_reclass_94[np.where(CV_reclass_94< 0.94)] = 0\n",
    "\n",
    "CV_reclass_95 = np.copy(CV)\n",
    "CV_reclass_95[np.where(CV_reclass_95>= 0.95)] = 1\n",
    "CV_reclass_95[np.where(CV_reclass_95< 0.95)] = 0\n",
    "\n",
    "CV_reclass_96 = np.copy(CV)\n",
    "CV_reclass_96[np.where(CV_reclass_96>= 0.96)] = 1\n",
    "CV_reclass_96[np.where(CV_reclass_96< 0.96)] = 0\n",
    "\n",
    "CV_reclass_97 = np.copy(CV)\n",
    "CV_reclass_97[np.where(CV_reclass_97>= 0.97)] = 1\n",
    "CV_reclass_97[np.where(CV_reclass_97< 0.97)] = 0\n",
    "\n",
    "CV_reclass_98 = np.copy(CV)\n",
    "CV_reclass_98[np.where(CV_reclass_98>= 0.98)] = 1\n",
    "CV_reclass_98[np.where(CV_reclass_98< 0.98)] = 0\n",
    "\n",
    "CV_reclass_99 = np.copy(CV)\n",
    "CV_reclass_99[np.where(CV_reclass_99>= 0.99)] = 1\n",
    "CV_reclass_99[np.where(CV_reclass_99< 0.99)] = 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Applying the watermask at each tested point. \n",
    "\n",
    "The watermask is applied by subtracting thewatermask from the CV reclassified raster \n",
    "\n",
    "        0-0 =     0     Non-crop\n",
    "        1-111 = -110    Water\n",
    "        1-0 =     1     Crop\n",
    "        0-111 = -111    Water\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#0.0 with water\n",
    "CV0_water = np.subtract(CV_reclass_0, watermask)\n",
    "CV01_water = np.subtract(CV_reclass_01, watermask)\n",
    "CV02_water = np.subtract(CV_reclass_02, watermask)\n",
    "CV03_water = np.subtract(CV_reclass_03, watermask)\n",
    "CV04_water = np.subtract(CV_reclass_04, watermask)\n",
    "CV05_water = np.subtract(CV_reclass_05, watermask)\n",
    "CV06_water = np.subtract(CV_reclass_06, watermask)\n",
    "CV07_water = np.subtract(CV_reclass_07, watermask)\n",
    "CV08_water = np.subtract(CV_reclass_08, watermask)\n",
    "CV09_water = np.subtract(CV_reclass_09, watermask)\n",
    "\n",
    "#0.1 with water\n",
    "CV1_water = np.subtract(CV_reclass_1, watermask)\n",
    "CV11_water = np.subtract(CV_reclass_11, watermask)\n",
    "CV12_water = np.subtract(CV_reclass_12, watermask)\n",
    "CV13_water = np.subtract(CV_reclass_13, watermask)\n",
    "CV14_water = np.subtract(CV_reclass_14, watermask)\n",
    "CV15_water = np.subtract(CV_reclass_15, watermask)\n",
    "CV16_water = np.subtract(CV_reclass_16, watermask)\n",
    "CV17_water = np.subtract(CV_reclass_17, watermask)\n",
    "CV18_water = np.subtract(CV_reclass_18, watermask)\n",
    "CV19_water = np.subtract(CV_reclass_19, watermask)\n",
    "\n",
    "#0.2 with water \n",
    "CV2_water = np.subtract(CV_reclass_2, watermask)\n",
    "CV21_water = np.subtract(CV_reclass_21, watermask)\n",
    "CV22_water = np.subtract(CV_reclass_22, watermask)\n",
    "CV23_water = np.subtract(CV_reclass_23, watermask)\n",
    "CV24_water = np.subtract(CV_reclass_24, watermask)\n",
    "CV25_water = np.subtract(CV_reclass_25, watermask)\n",
    "CV26_water = np.subtract(CV_reclass_26, watermask)\n",
    "CV27_water = np.subtract(CV_reclass_27, watermask)\n",
    "CV28_water = np.subtract(CV_reclass_28, watermask)\n",
    "CV29_water = np.subtract(CV_reclass_29, watermask)\n",
    "\n",
    "#0.3 with water\n",
    "CV3_water = np.subtract(CV_reclass_3, watermask)\n",
    "CV31_water = np.subtract(CV_reclass_31, watermask)\n",
    "CV32_water = np.subtract(CV_reclass_32, watermask)\n",
    "CV33_water = np.subtract(CV_reclass_33, watermask)\n",
    "CV34_water = np.subtract(CV_reclass_34, watermask)\n",
    "CV35_water = np.subtract(CV_reclass_35, watermask)\n",
    "CV36_water = np.subtract(CV_reclass_36, watermask)\n",
    "CV37_water = np.subtract(CV_reclass_37, watermask)\n",
    "CV38_water = np.subtract(CV_reclass_38, watermask)\n",
    "CV39_water = np.subtract(CV_reclass_39, watermask)\n",
    "\n",
    "#0.4 with water \n",
    "CV4_water = np.subtract(CV_reclass_4, watermask)\n",
    "CV41_water = np.subtract(CV_reclass_41, watermask)\n",
    "CV42_water = np.subtract(CV_reclass_42, watermask)\n",
    "CV43_water = np.subtract(CV_reclass_43, watermask)\n",
    "CV44_water = np.subtract(CV_reclass_44, watermask)\n",
    "CV45_water = np.subtract(CV_reclass_45, watermask)\n",
    "CV46_water = np.subtract(CV_reclass_46, watermask)\n",
    "CV47_water = np.subtract(CV_reclass_47, watermask)\n",
    "CV48_water = np.subtract(CV_reclass_48, watermask)\n",
    "CV49_water = np.subtract(CV_reclass_49, watermask)\n",
    "\n",
    "#0.5 with water \n",
    "CV5_water = np.subtract(CV_reclass_5, watermask)\n",
    "CV51_water = np.subtract(CV_reclass_51, watermask)\n",
    "CV52_water = np.subtract(CV_reclass_52, watermask)\n",
    "CV53_water = np.subtract(CV_reclass_53, watermask)\n",
    "CV54_water = np.subtract(CV_reclass_54, watermask)\n",
    "CV55_water = np.subtract(CV_reclass_55, watermask)\n",
    "CV56_water = np.subtract(CV_reclass_56, watermask)\n",
    "CV57_water = np.subtract(CV_reclass_57, watermask)\n",
    "CV58_water = np.subtract(CV_reclass_58, watermask)\n",
    "CV59_water = np.subtract(CV_reclass_59, watermask)\n",
    "\n",
    "#0.6 with water\n",
    "CV6_water = np.subtract(CV_reclass_6, watermask)\n",
    "CV61_water = np.subtract(CV_reclass_61, watermask)\n",
    "CV62_water = np.subtract(CV_reclass_62, watermask)\n",
    "CV63_water = np.subtract(CV_reclass_63, watermask)\n",
    "CV64_water = np.subtract(CV_reclass_64, watermask)\n",
    "CV65_water = np.subtract(CV_reclass_65, watermask)\n",
    "CV66_water = np.subtract(CV_reclass_66, watermask)\n",
    "CV67_water = np.subtract(CV_reclass_67, watermask)\n",
    "CV68_water = np.subtract(CV_reclass_68, watermask)\n",
    "CV69_water = np.subtract(CV_reclass_69, watermask)\n",
    "\n",
    "#0.7 with water \n",
    "CV7_water = np.subtract(CV_reclass_7, watermask)\n",
    "CV71_water = np.subtract(CV_reclass_71, watermask)\n",
    "CV72_water = np.subtract(CV_reclass_72, watermask)\n",
    "CV73_water = np.subtract(CV_reclass_73, watermask)\n",
    "CV74_water = np.subtract(CV_reclass_74, watermask)\n",
    "CV75_water = np.subtract(CV_reclass_75, watermask)\n",
    "CV76_water = np.subtract(CV_reclass_76, watermask)\n",
    "CV77_water = np.subtract(CV_reclass_77, watermask)\n",
    "CV78_water = np.subtract(CV_reclass_78, watermask)\n",
    "CV79_water = np.subtract(CV_reclass_79, watermask)\n",
    "\n",
    "#0.8 with water \n",
    "CV8_water = np.subtract(CV_reclass_8, watermask)\n",
    "CV81_water = np.subtract(CV_reclass_81, watermask)\n",
    "CV82_water = np.subtract(CV_reclass_82, watermask)\n",
    "CV83_water = np.subtract(CV_reclass_83, watermask)\n",
    "CV84_water = np.subtract(CV_reclass_84, watermask)\n",
    "CV85_water = np.subtract(CV_reclass_85, watermask)\n",
    "CV86_water = np.subtract(CV_reclass_86, watermask)\n",
    "CV87_water = np.subtract(CV_reclass_87, watermask)\n",
    "CV88_water = np.subtract(CV_reclass_88, watermask)\n",
    "CV89_water = np.subtract(CV_reclass_89, watermask)\n",
    "\n",
    "#0.9 with water \n",
    "CV9_water = np.subtract(CV_reclass_9, watermask)\n",
    "CV91_water = np.subtract(CV_reclass_91, watermask)\n",
    "CV92_water = np.subtract(CV_reclass_92, watermask)\n",
    "CV93_water = np.subtract(CV_reclass_93, watermask)\n",
    "CV94_water = np.subtract(CV_reclass_94, watermask)\n",
    "CV95_water = np.subtract(CV_reclass_95, watermask)\n",
    "CV96_water = np.subtract(CV_reclass_96, watermask)\n",
    "CV97_water = np.subtract(CV_reclass_97, watermask)\n",
    "CV98_water = np.subtract(CV_reclass_98, watermask)\n",
    "CV99_water = np.subtract(CV_reclass_99, watermask)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Creating a binary crop/non-crop classification at each tested threshold.\n",
    "\n",
    "          1 = crop\n",
    "        -10 = noncrop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#0.0 crop/non\n",
    "CV0_crop_non = np.copy(CV0_water)\n",
    "CV0_crop_non[np.where(CV0_crop_non>0)] = 1\n",
    "CV0_crop_non[np.where(CV0_crop_non<=0)] = -10\n",
    "\n",
    "CV01_crop_non = np.copy(CV01_water)\n",
    "CV01_crop_non[np.where(CV01_crop_non>0)] = 1\n",
    "CV01_crop_non[np.where(CV01_crop_non<=0)] = -10\n",
    "\n",
    "CV02_crop_non = np.copy(CV02_water)\n",
    "CV02_crop_non[np.where(CV02_crop_non>0)] = 1\n",
    "CV02_crop_non[np.where(CV02_crop_non<=0)] = -10\n",
    "\n",
    "CV03_crop_non = np.copy(CV03_water)\n",
    "CV03_crop_non[np.where(CV03_crop_non>0)] = 1\n",
    "CV03_crop_non[np.where(CV03_crop_non<=0)] = -10\n",
    "\n",
    "CV04_crop_non = np.copy(CV04_water)\n",
    "CV04_crop_non[np.where(CV04_crop_non>0)] = 1\n",
    "CV04_crop_non[np.where(CV04_crop_non<=0)] = -10\n",
    "\n",
    "CV05_crop_non = np.copy(CV05_water)\n",
    "CV05_crop_non[np.where(CV05_crop_non>0)] = 1\n",
    "CV05_crop_non[np.where(CV05_crop_non<=0)] = -10\n",
    "\n",
    "CV06_crop_non = np.copy(CV06_water)\n",
    "CV06_crop_non[np.where(CV06_crop_non>0)] = 1\n",
    "CV06_crop_non[np.where(CV06_crop_non<=0)] = -10\n",
    "\n",
    "CV07_crop_non = np.copy(CV07_water)\n",
    "CV07_crop_non[np.where(CV07_crop_non>0)] = 1\n",
    "CV07_crop_non[np.where(CV07_crop_non<=0)] = -10\n",
    "\n",
    "CV08_crop_non = np.copy(CV08_water)\n",
    "CV08_crop_non[np.where(CV08_crop_non>0)] = 1\n",
    "CV08_crop_non[np.where(CV08_crop_non<=0)] = -10\n",
    "\n",
    "CV09_crop_non = np.copy(CV09_water)\n",
    "CV09_crop_non[np.where(CV09_crop_non>0)] = 1\n",
    "CV09_crop_non[np.where(CV09_crop_non<=0)] = -10\n",
    "\n",
    "#0.1 crop/non\n",
    "CV1_crop_non = np.copy(CV1_water)\n",
    "CV1_crop_non[np.where(CV1_crop_non>0)] = 1\n",
    "CV1_crop_non[np.where(CV1_crop_non<=0)] = -10\n",
    "\n",
    "CV11_crop_non = np.copy(CV11_water)\n",
    "CV11_crop_non[np.where(CV11_crop_non>0)] = 1\n",
    "CV11_crop_non[np.where(CV11_crop_non<=0)] = -10\n",
    "\n",
    "CV12_crop_non = np.copy(CV12_water)\n",
    "CV12_crop_non[np.where(CV12_crop_non>0)] = 1\n",
    "CV12_crop_non[np.where(CV12_crop_non<=0)] = -10\n",
    "\n",
    "CV13_crop_non = np.copy(CV13_water)\n",
    "CV13_crop_non[np.where(CV13_crop_non>0)] = 1\n",
    "CV13_crop_non[np.where(CV13_crop_non<=0)] = -10\n",
    "\n",
    "CV14_crop_non = np.copy(CV14_water)\n",
    "CV14_crop_non[np.where(CV14_crop_non>0)] = 1\n",
    "CV14_crop_non[np.where(CV14_crop_non<=0)] = -10\n",
    "\n",
    "CV15_crop_non = np.copy(CV15_water)\n",
    "CV15_crop_non[np.where(CV15_crop_non>0)] = 1\n",
    "CV15_crop_non[np.where(CV15_crop_non<=0)] = -10\n",
    "\n",
    "CV16_crop_non = np.copy(CV16_water)\n",
    "CV16_crop_non[np.where(CV16_crop_non>0)] = 1\n",
    "CV16_crop_non[np.where(CV16_crop_non<=0)] = -10\n",
    "\n",
    "CV17_crop_non = np.copy(CV17_water)\n",
    "CV17_crop_non[np.where(CV17_crop_non>0)] = 1\n",
    "CV17_crop_non[np.where(CV17_crop_non<=0)] = -10\n",
    "\n",
    "CV18_crop_non = np.copy(CV18_water)\n",
    "CV18_crop_non[np.where(CV18_crop_non>0)] = 1\n",
    "CV18_crop_non[np.where(CV18_crop_non<=0)] = -10\n",
    "\n",
    "CV19_crop_non = np.copy(CV19_water)\n",
    "CV19_crop_non[np.where(CV19_crop_non>0)] = 1\n",
    "CV19_crop_non[np.where(CV19_crop_non<=0)] = -10\n",
    "\n",
    "#0.2 crop/non\n",
    "CV2_crop_non = np.copy(CV2_water)\n",
    "CV2_crop_non[np.where(CV2_crop_non>0)] = 1\n",
    "CV2_crop_non[np.where(CV2_crop_non<=0)] = -10\n",
    "\n",
    "CV21_crop_non = np.copy(CV21_water)\n",
    "CV21_crop_non[np.where(CV21_crop_non>0)] = 1\n",
    "CV21_crop_non[np.where(CV21_crop_non<=0)] = -10\n",
    "\n",
    "CV22_crop_non = np.copy(CV22_water)\n",
    "CV22_crop_non[np.where(CV22_crop_non>0)] = 1\n",
    "CV22_crop_non[np.where(CV22_crop_non<=0)] = -10\n",
    "\n",
    "CV23_crop_non = np.copy(CV23_water)\n",
    "CV23_crop_non[np.where(CV23_crop_non>0)] = 1\n",
    "CV23_crop_non[np.where(CV23_crop_non<=0)] = -10\n",
    "\n",
    "CV24_crop_non = np.copy(CV24_water)\n",
    "CV24_crop_non[np.where(CV24_crop_non>0)] = 1\n",
    "CV24_crop_non[np.where(CV24_crop_non<=0)] = -10\n",
    "\n",
    "CV25_crop_non = np.copy(CV25_water)\n",
    "CV25_crop_non[np.where(CV25_crop_non>0)] = 1\n",
    "CV25_crop_non[np.where(CV25_crop_non<=0)] = -10\n",
    "\n",
    "CV26_crop_non = np.copy(CV26_water)\n",
    "CV26_crop_non[np.where(CV26_crop_non>0)] = 1\n",
    "CV26_crop_non[np.where(CV26_crop_non<=0)] = -10\n",
    "\n",
    "CV27_crop_non = np.copy(CV27_water)\n",
    "CV27_crop_non[np.where(CV27_crop_non>0)] = 1\n",
    "CV27_crop_non[np.where(CV27_crop_non<=0)] = -10\n",
    "\n",
    "CV28_crop_non = np.copy(CV28_water)\n",
    "CV28_crop_non[np.where(CV28_crop_non>0)] = 1\n",
    "CV28_crop_non[np.where(CV28_crop_non<=0)] = -10\n",
    "\n",
    "CV29_crop_non = np.copy(CV29_water)\n",
    "CV29_crop_non[np.where(CV29_crop_non>0)] = 1\n",
    "CV29_crop_non[np.where(CV29_crop_non<=0)] = -10\n",
    "\n",
    "#0.3 crop/non\n",
    "CV3_crop_non = np.copy(CV3_water)\n",
    "CV3_crop_non[np.where(CV3_crop_non>0)] = 1\n",
    "CV3_crop_non[np.where(CV3_crop_non<=0)] = -10\n",
    "\n",
    "CV31_crop_non = np.copy(CV31_water)\n",
    "CV31_crop_non[np.where(CV31_crop_non>0)] = 1\n",
    "CV31_crop_non[np.where(CV31_crop_non<=0)] = -10\n",
    "\n",
    "CV32_crop_non = np.copy(CV32_water)\n",
    "CV32_crop_non[np.where(CV32_crop_non>0)] = 1\n",
    "CV32_crop_non[np.where(CV32_crop_non<=0)] = -10\n",
    "\n",
    "CV33_crop_non = np.copy(CV33_water)\n",
    "CV33_crop_non[np.where(CV33_crop_non>0)] = 1\n",
    "CV33_crop_non[np.where(CV33_crop_non<=0)] = -10\n",
    "\n",
    "CV34_crop_non = np.copy(CV34_water)\n",
    "CV34_crop_non[np.where(CV34_crop_non>0)] = 1\n",
    "CV34_crop_non[np.where(CV34_crop_non<=0)] = -10\n",
    "\n",
    "CV35_crop_non = np.copy(CV35_water)\n",
    "CV35_crop_non[np.where(CV35_crop_non>0)] = 1\n",
    "CV35_crop_non[np.where(CV35_crop_non<=0)] = -10\n",
    "\n",
    "CV36_crop_non = np.copy(CV36_water)\n",
    "CV36_crop_non[np.where(CV36_crop_non>0)] = 1\n",
    "CV36_crop_non[np.where(CV36_crop_non<=0)] = -10\n",
    "\n",
    "CV37_crop_non = np.copy(CV37_water)\n",
    "CV37_crop_non[np.where(CV37_crop_non>0)] = 1\n",
    "CV37_crop_non[np.where(CV37_crop_non<=0)] = -10\n",
    "\n",
    "CV38_crop_non = np.copy(CV38_water)\n",
    "CV38_crop_non[np.where(CV38_crop_non>0)] = 1\n",
    "CV38_crop_non[np.where(CV38_crop_non<=0)] = -10\n",
    "\n",
    "CV39_crop_non = np.copy(CV39_water)\n",
    "CV39_crop_non[np.where(CV39_crop_non>0)] = 1\n",
    "CV39_crop_non[np.where(CV39_crop_non<=0)] = -10\n",
    "\n",
    "#0.4 crop/non\n",
    "CV4_crop_non = np.copy(CV4_water)\n",
    "CV4_crop_non[np.where(CV4_crop_non>0)] = 1\n",
    "CV4_crop_non[np.where(CV4_crop_non<=0)] = -10\n",
    "\n",
    "CV41_crop_non = np.copy(CV41_water)\n",
    "CV41_crop_non[np.where(CV41_crop_non>0)] = 1\n",
    "CV41_crop_non[np.where(CV41_crop_non<=0)] = -10\n",
    "\n",
    "CV42_crop_non = np.copy(CV42_water)\n",
    "CV42_crop_non[np.where(CV42_crop_non>0)] = 1\n",
    "CV42_crop_non[np.where(CV42_crop_non<=0)] = -10\n",
    "\n",
    "CV43_crop_non = np.copy(CV43_water)\n",
    "CV43_crop_non[np.where(CV43_crop_non>0)] = 1\n",
    "CV43_crop_non[np.where(CV43_crop_non<=0)] = -10\n",
    "\n",
    "CV44_crop_non = np.copy(CV44_water)\n",
    "CV44_crop_non[np.where(CV44_crop_non>0)] = 1\n",
    "CV44_crop_non[np.where(CV44_crop_non<=0)] = -10\n",
    "\n",
    "CV45_crop_non = np.copy(CV45_water)\n",
    "CV45_crop_non[np.where(CV45_crop_non>0)] = 1\n",
    "CV45_crop_non[np.where(CV45_crop_non<=0)] = -10\n",
    "\n",
    "CV46_crop_non = np.copy(CV46_water)\n",
    "CV46_crop_non[np.where(CV46_crop_non>0)] = 1\n",
    "CV46_crop_non[np.where(CV46_crop_non<=0)] = -10\n",
    "\n",
    "CV47_crop_non = np.copy(CV47_water)\n",
    "CV47_crop_non[np.where(CV47_crop_non>0)] = 1\n",
    "CV47_crop_non[np.where(CV47_crop_non<=0)] = -10\n",
    "\n",
    "CV48_crop_non = np.copy(CV48_water)\n",
    "CV48_crop_non[np.where(CV48_crop_non>0)] = 1\n",
    "CV48_crop_non[np.where(CV48_crop_non<=0)] = -10\n",
    "\n",
    "CV49_crop_non = np.copy(CV49_water)\n",
    "CV49_crop_non[np.where(CV49_crop_non>0)] = 1\n",
    "CV49_crop_non[np.where(CV49_crop_non<=0)] = -10\n",
    "\n",
    "#0.5 crop/non\n",
    "CV5_crop_non = np.copy(CV5_water)\n",
    "CV5_crop_non[np.where(CV5_crop_non>0)] = 1\n",
    "CV5_crop_non[np.where(CV5_crop_non<=0)] = -10\n",
    "\n",
    "CV51_crop_non = np.copy(CV51_water)\n",
    "CV51_crop_non[np.where(CV51_crop_non>0)] = 1\n",
    "CV51_crop_non[np.where(CV51_crop_non<=0)] = -10\n",
    "\n",
    "CV52_crop_non = np.copy(CV52_water)\n",
    "CV52_crop_non[np.where(CV52_crop_non>0)] = 1\n",
    "CV52_crop_non[np.where(CV52_crop_non<=0)] = -10\n",
    "\n",
    "CV53_crop_non = np.copy(CV53_water)\n",
    "CV53_crop_non[np.where(CV53_crop_non>0)] = 1\n",
    "CV53_crop_non[np.where(CV53_crop_non<=0)] = -10\n",
    "\n",
    "CV54_crop_non = np.copy(CV54_water)\n",
    "CV54_crop_non[np.where(CV54_crop_non>0)] = 1\n",
    "CV54_crop_non[np.where(CV54_crop_non<=0)] = -10\n",
    "\n",
    "CV55_crop_non = np.copy(CV55_water)\n",
    "CV55_crop_non[np.where(CV55_crop_non>0)] = 1\n",
    "CV55_crop_non[np.where(CV55_crop_non<=0)] = -10\n",
    "\n",
    "CV56_crop_non = np.copy(CV56_water)\n",
    "CV56_crop_non[np.where(CV56_crop_non>0)] = 1\n",
    "CV56_crop_non[np.where(CV56_crop_non<=0)] = -10\n",
    "\n",
    "CV57_crop_non = np.copy(CV57_water)\n",
    "CV57_crop_non[np.where(CV57_crop_non>0)] = 1\n",
    "CV57_crop_non[np.where(CV57_crop_non<=0)] = -10\n",
    "\n",
    "CV58_crop_non = np.copy(CV58_water)\n",
    "CV58_crop_non[np.where(CV58_crop_non>0)] = 1\n",
    "CV58_crop_non[np.where(CV58_crop_non<=0)] = -10\n",
    "\n",
    "CV59_crop_non = np.copy(CV59_water)\n",
    "CV59_crop_non[np.where(CV59_crop_non>0)] = 1\n",
    "CV59_crop_non[np.where(CV59_crop_non<=0)] = -10\n",
    "\n",
    "#0.6 crop/non\n",
    "CV6_crop_non = np.copy(CV6_water)\n",
    "CV6_crop_non[np.where(CV6_crop_non>0)] = 1\n",
    "CV6_crop_non[np.where(CV6_crop_non<=0)] = -10\n",
    "\n",
    "CV61_crop_non = np.copy(CV61_water)\n",
    "CV61_crop_non[np.where(CV61_crop_non>0)] = 1\n",
    "CV61_crop_non[np.where(CV61_crop_non<=0)] = -10\n",
    "\n",
    "CV62_crop_non = np.copy(CV62_water)\n",
    "CV62_crop_non[np.where(CV62_crop_non>0)] = 1\n",
    "CV62_crop_non[np.where(CV62_crop_non<=0)] = -10\n",
    "\n",
    "CV63_crop_non = np.copy(CV63_water)\n",
    "CV63_crop_non[np.where(CV63_crop_non>0)] = 1\n",
    "CV63_crop_non[np.where(CV63_crop_non<=0)] = -10\n",
    "\n",
    "CV64_crop_non = np.copy(CV64_water)\n",
    "CV64_crop_non[np.where(CV64_crop_non>0)] = 1\n",
    "CV64_crop_non[np.where(CV64_crop_non<=0)] = -10\n",
    "\n",
    "CV65_crop_non = np.copy(CV65_water)\n",
    "CV65_crop_non[np.where(CV65_crop_non>0)] = 1\n",
    "CV65_crop_non[np.where(CV65_crop_non<=0)] = -10\n",
    "\n",
    "CV66_crop_non = np.copy(CV66_water)\n",
    "CV66_crop_non[np.where(CV66_crop_non>0)] = 1\n",
    "CV66_crop_non[np.where(CV66_crop_non<=0)] = -10\n",
    "\n",
    "CV67_crop_non = np.copy(CV67_water)\n",
    "CV67_crop_non[np.where(CV67_crop_non>0)] = 1\n",
    "CV67_crop_non[np.where(CV67_crop_non<=0)] = -10\n",
    "\n",
    "CV68_crop_non = np.copy(CV68_water)\n",
    "CV68_crop_non[np.where(CV68_crop_non>0)] = 1\n",
    "CV68_crop_non[np.where(CV68_crop_non<=0)] = -10\n",
    "\n",
    "CV69_crop_non = np.copy(CV69_water)\n",
    "CV69_crop_non[np.where(CV69_crop_non>0)] = 1\n",
    "CV69_crop_non[np.where(CV69_crop_non<=0)] = -10\n",
    "\n",
    "#0.7 crop/non\n",
    "CV7_crop_non = np.copy(CV7_water)\n",
    "CV7_crop_non[np.where(CV7_crop_non>0)] = 1\n",
    "CV7_crop_non[np.where(CV7_crop_non<=0)] = -10\n",
    "\n",
    "CV71_crop_non = np.copy(CV71_water)\n",
    "CV71_crop_non[np.where(CV71_crop_non>0)] = 1\n",
    "CV71_crop_non[np.where(CV71_crop_non<=0)] = -10\n",
    "\n",
    "CV72_crop_non = np.copy(CV72_water)\n",
    "CV72_crop_non[np.where(CV72_crop_non>0)] = 1\n",
    "CV72_crop_non[np.where(CV72_crop_non<=0)] = -10\n",
    "\n",
    "CV73_crop_non = np.copy(CV73_water)\n",
    "CV73_crop_non[np.where(CV73_crop_non>0)] = 1\n",
    "CV73_crop_non[np.where(CV73_crop_non<=0)] = -10\n",
    "\n",
    "CV74_crop_non = np.copy(CV74_water)\n",
    "CV74_crop_non[np.where(CV74_crop_non>0)] = 1\n",
    "CV74_crop_non[np.where(CV74_crop_non<=0)] = -10\n",
    "\n",
    "CV75_crop_non = np.copy(CV75_water)\n",
    "CV75_crop_non[np.where(CV75_crop_non>0)] = 1\n",
    "CV75_crop_non[np.where(CV75_crop_non<=0)] = -10\n",
    "\n",
    "CV76_crop_non = np.copy(CV76_water)\n",
    "CV76_crop_non[np.where(CV76_crop_non>0)] = 1\n",
    "CV76_crop_non[np.where(CV76_crop_non<=0)] = -10\n",
    "\n",
    "CV77_crop_non = np.copy(CV77_water)\n",
    "CV77_crop_non[np.where(CV77_crop_non>0)] = 1\n",
    "CV77_crop_non[np.where(CV77_crop_non<=0)] = -10\n",
    "\n",
    "CV78_crop_non = np.copy(CV78_water)\n",
    "CV78_crop_non[np.where(CV78_crop_non>0)] = 1\n",
    "CV78_crop_non[np.where(CV78_crop_non<=0)] = -10\n",
    "\n",
    "CV79_crop_non = np.copy(CV79_water)\n",
    "CV79_crop_non[np.where(CV79_crop_non>0)] = 1\n",
    "CV79_crop_non[np.where(CV79_crop_non<=0)] = -10\n",
    "\n",
    "#0.8 crop/non\n",
    "CV8_crop_non = np.copy(CV8_water)\n",
    "CV8_crop_non[np.where(CV8_crop_non>0)] = 1\n",
    "CV8_crop_non[np.where(CV8_crop_non<=0)] = -10\n",
    "\n",
    "CV81_crop_non = np.copy(CV81_water)\n",
    "CV81_crop_non[np.where(CV81_crop_non>0)] = 1\n",
    "CV81_crop_non[np.where(CV81_crop_non<=0)] = -10\n",
    "\n",
    "CV82_crop_non = np.copy(CV82_water)\n",
    "CV82_crop_non[np.where(CV82_crop_non>0)] = 1\n",
    "CV82_crop_non[np.where(CV82_crop_non<=0)] = -10\n",
    "\n",
    "CV83_crop_non = np.copy(CV83_water)\n",
    "CV83_crop_non[np.where(CV83_crop_non>0)] = 1\n",
    "CV83_crop_non[np.where(CV83_crop_non<=0)] = -10\n",
    "\n",
    "CV84_crop_non = np.copy(CV84_water)\n",
    "CV84_crop_non[np.where(CV84_crop_non>0)] = 1\n",
    "CV84_crop_non[np.where(CV84_crop_non<=0)] = -10\n",
    "\n",
    "CV85_crop_non = np.copy(CV85_water)\n",
    "CV85_crop_non[np.where(CV85_crop_non>0)] = 1\n",
    "CV85_crop_non[np.where(CV85_crop_non<=0)] = -10\n",
    "\n",
    "CV86_crop_non = np.copy(CV86_water)\n",
    "CV86_crop_non[np.where(CV86_crop_non>0)] = 1\n",
    "CV86_crop_non[np.where(CV86_crop_non<=0)] = -10\n",
    "\n",
    "CV87_crop_non = np.copy(CV87_water)\n",
    "CV87_crop_non[np.where(CV87_crop_non>0)] = 1\n",
    "CV87_crop_non[np.where(CV87_crop_non<=0)] = -10\n",
    "\n",
    "CV88_crop_non = np.copy(CV88_water)\n",
    "CV88_crop_non[np.where(CV88_crop_non>0)] = 1\n",
    "CV88_crop_non[np.where(CV88_crop_non<=0)] = -10\n",
    "\n",
    "CV89_crop_non = np.copy(CV89_water)\n",
    "CV89_crop_non[np.where(CV89_crop_non>0)] = 1\n",
    "CV89_crop_non[np.where(CV89_crop_non<=0)] = -10\n",
    "\n",
    "#0.9 crop/non\n",
    "CV9_crop_non = np.copy(CV9_water)\n",
    "CV9_crop_non[np.where(CV9_crop_non>0)] = 1\n",
    "CV9_crop_non[np.where(CV9_crop_non<=0)] = -10\n",
    "\n",
    "CV91_crop_non = np.copy(CV91_water)\n",
    "CV91_crop_non[np.where(CV91_crop_non>0)] = 1\n",
    "CV91_crop_non[np.where(CV91_crop_non<=0)] = -10\n",
    "\n",
    "CV92_crop_non = np.copy(CV92_water)\n",
    "CV92_crop_non[np.where(CV92_crop_non>0)] = 1\n",
    "CV92_crop_non[np.where(CV92_crop_non<=0)] = -10\n",
    "\n",
    "CV93_crop_non = np.copy(CV93_water)\n",
    "CV93_crop_non[np.where(CV93_crop_non>0)] = 1\n",
    "CV93_crop_non[np.where(CV93_crop_non<=0)] = -10\n",
    "\n",
    "CV94_crop_non = np.copy(CV94_water)\n",
    "CV94_crop_non[np.where(CV94_crop_non>0)] = 1\n",
    "CV94_crop_non[np.where(CV94_crop_non<=0)] = -10\n",
    "\n",
    "CV95_crop_non = np.copy(CV95_water)\n",
    "CV95_crop_non[np.where(CV95_crop_non>0)] = 1\n",
    "CV95_crop_non[np.where(CV95_crop_non<=0)] = -10\n",
    "\n",
    "CV96_crop_non = np.copy(CV96_water)\n",
    "CV96_crop_non[np.where(CV96_crop_non>0)] = 1\n",
    "CV96_crop_non[np.where(CV96_crop_non<=0)] = -10\n",
    "\n",
    "CV97_crop_non = np.copy(CV97_water)\n",
    "CV97_crop_non[np.where(CV97_crop_non>0)] = 1\n",
    "CV97_crop_non[np.where(CV97_crop_non<=0)] = -10\n",
    "\n",
    "CV98_crop_non = np.copy(CV98_water)\n",
    "CV98_crop_non[np.where(CV98_crop_non>0)] = 1\n",
    "CV98_crop_non[np.where(CV98_crop_non<=0)] = -10\n",
    "\n",
    "CV99_crop_non = np.copy(CV99_water)\n",
    "CV99_crop_non[np.where(CV99_crop_non>0)] = 1\n",
    "CV99_crop_non[np.where(CV99_crop_non<=0)] = -10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Finding the correctly and incorrectly classified pixels from each classification.\n",
    "\n",
    "Subtracted the binary crop/non-crop classification from the correct binary crop/non-crop classification from the CDL.\n",
    "\n",
    "        0 - -10 = 10       non-crop correctly classified as non-crop\n",
    "        1 - 1 =   0        crop correctly classified as crop\n",
    "        0 - 1 =  -1        non-crop incorrectly classified as crop\n",
    "        1- -10 =  11       crop incorrectly classified as non-crop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#0.0 with CDL\n",
    "CDL_CV0 = np.subtract(CDL_binary, CV0_crop_non)\n",
    "CDL_CV01 = np.subtract(CDL_binary, CV01_crop_non)\n",
    "CDL_CV02 = np.subtract(CDL_binary, CV02_crop_non)\n",
    "CDL_CV03 = np.subtract(CDL_binary, CV03_crop_non)\n",
    "CDL_CV04 = np.subtract(CDL_binary, CV04_crop_non)\n",
    "CDL_CV05 = np.subtract(CDL_binary, CV05_crop_non)\n",
    "CDL_CV06 = np.subtract(CDL_binary, CV06_crop_non)\n",
    "CDL_CV07 = np.subtract(CDL_binary, CV07_crop_non)\n",
    "CDL_CV08 = np.subtract(CDL_binary, CV08_crop_non)\n",
    "CDL_CV09 = np.subtract(CDL_binary, CV09_crop_non)\n",
    "\n",
    "#0.1 with CDL\n",
    "CDL_CV1 = np.subtract(CDL_binary, CV1_crop_non)\n",
    "CDL_CV11 = np.subtract(CDL_binary, CV11_crop_non)\n",
    "CDL_CV12 = np.subtract(CDL_binary, CV12_crop_non)\n",
    "CDL_CV13 = np.subtract(CDL_binary, CV13_crop_non)\n",
    "CDL_CV14 = np.subtract(CDL_binary, CV14_crop_non)\n",
    "CDL_CV15 = np.subtract(CDL_binary, CV15_crop_non)\n",
    "CDL_CV16 = np.subtract(CDL_binary, CV16_crop_non)\n",
    "CDL_CV17 = np.subtract(CDL_binary, CV17_crop_non)\n",
    "CDL_CV18 = np.subtract(CDL_binary, CV18_crop_non)\n",
    "CDL_CV19 = np.subtract(CDL_binary, CV19_crop_non)\n",
    "\n",
    "#0.2 with CDL\n",
    "CDL_CV2 = np.subtract(CDL_binary, CV2_crop_non)\n",
    "CDL_CV21 = np.subtract(CDL_binary, CV21_crop_non)\n",
    "CDL_CV22 = np.subtract(CDL_binary, CV22_crop_non)\n",
    "CDL_CV23 = np.subtract(CDL_binary, CV23_crop_non)\n",
    "CDL_CV24 = np.subtract(CDL_binary, CV24_crop_non)\n",
    "CDL_CV25 = np.subtract(CDL_binary, CV25_crop_non)\n",
    "CDL_CV26 = np.subtract(CDL_binary, CV26_crop_non)\n",
    "CDL_CV27 = np.subtract(CDL_binary, CV27_crop_non)\n",
    "CDL_CV28 = np.subtract(CDL_binary, CV28_crop_non)\n",
    "CDL_CV29 = np.subtract(CDL_binary, CV29_crop_non)\n",
    "\n",
    "#0.3 with CDL\n",
    "CDL_CV3 = np.subtract(CDL_binary, CV3_crop_non)\n",
    "CDL_CV31 = np.subtract(CDL_binary, CV31_crop_non)\n",
    "CDL_CV32 = np.subtract(CDL_binary, CV32_crop_non)\n",
    "CDL_CV33 = np.subtract(CDL_binary, CV33_crop_non)\n",
    "CDL_CV34 = np.subtract(CDL_binary, CV34_crop_non)\n",
    "CDL_CV35 = np.subtract(CDL_binary, CV35_crop_non)\n",
    "CDL_CV36 = np.subtract(CDL_binary, CV36_crop_non)\n",
    "CDL_CV37 = np.subtract(CDL_binary, CV37_crop_non)\n",
    "CDL_CV38 = np.subtract(CDL_binary, CV38_crop_non)\n",
    "CDL_CV39 = np.subtract(CDL_binary, CV39_crop_non)\n",
    "\n",
    "#0.4 with CDL\n",
    "CDL_CV4 = np.subtract(CDL_binary, CV4_crop_non)\n",
    "CDL_CV41 = np.subtract(CDL_binary, CV41_crop_non)\n",
    "CDL_CV42 = np.subtract(CDL_binary, CV42_crop_non)\n",
    "CDL_CV43 = np.subtract(CDL_binary, CV43_crop_non)\n",
    "CDL_CV44 = np.subtract(CDL_binary, CV44_crop_non)\n",
    "CDL_CV45 = np.subtract(CDL_binary, CV45_crop_non)\n",
    "CDL_CV46 = np.subtract(CDL_binary, CV46_crop_non)\n",
    "CDL_CV47 = np.subtract(CDL_binary, CV47_crop_non)\n",
    "CDL_CV48 = np.subtract(CDL_binary, CV48_crop_non)\n",
    "CDL_CV49 = np.subtract(CDL_binary, CV49_crop_non)\n",
    "\n",
    "#0.5 with CDL\n",
    "CDL_CV5 = np.subtract(CDL_binary, CV5_crop_non)\n",
    "CDL_CV51 = np.subtract(CDL_binary, CV51_crop_non)\n",
    "CDL_CV52 = np.subtract(CDL_binary, CV52_crop_non)\n",
    "CDL_CV53 = np.subtract(CDL_binary, CV53_crop_non)\n",
    "CDL_CV54 = np.subtract(CDL_binary, CV54_crop_non)\n",
    "CDL_CV55 = np.subtract(CDL_binary, CV55_crop_non)\n",
    "CDL_CV56 = np.subtract(CDL_binary, CV56_crop_non)\n",
    "CDL_CV57 = np.subtract(CDL_binary, CV57_crop_non)\n",
    "CDL_CV58 = np.subtract(CDL_binary, CV58_crop_non)\n",
    "CDL_CV59 = np.subtract(CDL_binary, CV59_crop_non)\n",
    "\n",
    "#0.6 with CDL\n",
    "CDL_CV6 = np.subtract(CDL_binary, CV6_crop_non)\n",
    "CDL_CV61 = np.subtract(CDL_binary, CV61_crop_non)\n",
    "CDL_CV62 = np.subtract(CDL_binary, CV62_crop_non)\n",
    "CDL_CV63 = np.subtract(CDL_binary, CV63_crop_non)\n",
    "CDL_CV64 = np.subtract(CDL_binary, CV64_crop_non)\n",
    "CDL_CV65 = np.subtract(CDL_binary, CV65_crop_non)\n",
    "CDL_CV66 = np.subtract(CDL_binary, CV66_crop_non)\n",
    "CDL_CV67 = np.subtract(CDL_binary, CV67_crop_non)\n",
    "CDL_CV68 = np.subtract(CDL_binary, CV68_crop_non)\n",
    "CDL_CV69 = np.subtract(CDL_binary, CV69_crop_non)\n",
    "\n",
    "#0.7 with CDL\n",
    "CDL_CV7 = np.subtract(CDL_binary, CV7_crop_non)\n",
    "CDL_CV71 = np.subtract(CDL_binary, CV71_crop_non)\n",
    "CDL_CV72 = np.subtract(CDL_binary, CV72_crop_non)\n",
    "CDL_CV73 = np.subtract(CDL_binary, CV73_crop_non)\n",
    "CDL_CV74 = np.subtract(CDL_binary, CV74_crop_non)\n",
    "CDL_CV75 = np.subtract(CDL_binary, CV75_crop_non)\n",
    "CDL_CV76 = np.subtract(CDL_binary, CV76_crop_non)\n",
    "CDL_CV77 = np.subtract(CDL_binary, CV77_crop_non)\n",
    "CDL_CV78 = np.subtract(CDL_binary, CV78_crop_non)\n",
    "CDL_CV79 = np.subtract(CDL_binary, CV79_crop_non)\n",
    "\n",
    "#0.8 with CDL\n",
    "CDL_CV8 = np.subtract(CDL_binary, CV8_crop_non)\n",
    "CDL_CV81 = np.subtract(CDL_binary, CV81_crop_non)\n",
    "CDL_CV82 = np.subtract(CDL_binary, CV82_crop_non)\n",
    "CDL_CV83 = np.subtract(CDL_binary, CV83_crop_non)\n",
    "CDL_CV84 = np.subtract(CDL_binary, CV84_crop_non)\n",
    "CDL_CV85 = np.subtract(CDL_binary, CV85_crop_non)\n",
    "CDL_CV86 = np.subtract(CDL_binary, CV86_crop_non)\n",
    "CDL_CV87 = np.subtract(CDL_binary, CV87_crop_non)\n",
    "CDL_CV88 = np.subtract(CDL_binary, CV88_crop_non)\n",
    "CDL_CV89 = np.subtract(CDL_binary, CV89_crop_non)\n",
    "\n",
    "#0.9 with CDL\n",
    "CDL_CV9 = np.subtract(CDL_binary, CV9_crop_non)\n",
    "CDL_CV91 = np.subtract(CDL_binary, CV91_crop_non)\n",
    "CDL_CV92 = np.subtract(CDL_binary, CV92_crop_non)\n",
    "CDL_CV93 = np.subtract(CDL_binary, CV93_crop_non)\n",
    "CDL_CV94 = np.subtract(CDL_binary, CV94_crop_non)\n",
    "CDL_CV95 = np.subtract(CDL_binary, CV95_crop_non)\n",
    "CDL_CV96 = np.subtract(CDL_binary, CV96_crop_non)\n",
    "CDL_CV97 = np.subtract(CDL_binary, CV97_crop_non)\n",
    "CDL_CV98 = np.subtract(CDL_binary, CV98_crop_non)\n",
    "CDL_CV99 = np.subtract(CDL_binary, CV99_crop_non)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Calculating sensitivity and specificity:\n",
    "\n",
    "True Positive (TP) = Crop correctly classified as crop\n",
    "\n",
    "False Positive (FP) = Crop incorrectly classified as non-crop\n",
    "\n",
    "True Negitive (TN) = Non-crop correctly classified as non-crop\n",
    "\n",
    "False Negative (FN) = Non-crop incorrectly classified as crop\n",
    "\n",
    "True Positive Rate (TPR) = Sensitivity\n",
    "\n",
    "False Positive Rate (FPR) = Specificity\n",
    "\n",
    "Sensitivity = TP / (TP + TN)\n",
    "\n",
    "Specificity = TN / (TN + FP)\n",
    "\n",
    "For use in the generation of a ROC curve (1 - Specificity) is plotted by Sensitivity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#0.0 sensitivity and specificity \n",
    "denom_sens_0 = np.add(np.count_nonzero(CDL_CV0 == 0), np.count_nonzero(CDL_CV0 == 11))\n",
    "sensitivity0_y = np.divide(np.count_nonzero(CDL_CV0 == 0),denom_sens_0)\n",
    "denom_spec_0 = np.add(np.count_nonzero(CDL_CV0 == 10)- cdl111, np.count_nonzero(CDL_CV0 == -1))\n",
    "specificity0_x = np.divide(np.count_nonzero(CDL_CV0 == 10)- cdl111,denom_spec_0)\n",
    "sub_specificity0_x = np.subtract(1, specificity0_x)\n",
    "\n",
    "denom_sens_01 = np.add(np.count_nonzero(CDL_CV01 == 0), np.count_nonzero(CDL_CV01 == 11))\n",
    "sensitivity01_y = np.divide(np.count_nonzero(CDL_CV01 == 0),denom_sens_01)\n",
    "denom_spec_01 = np.add(np.count_nonzero(CDL_CV01 == 10)- cdl111, np.count_nonzero(CDL_CV01 == -1))\n",
    "specificity01_x = np.divide(np.count_nonzero(CDL_CV01 == 10)- cdl111,denom_spec_01)\n",
    "sub_specificity01_x = np.subtract(1, specificity01_x)\n",
    "\n",
    "denom_sens_02 = np.add(np.count_nonzero(CDL_CV02 == 0), np.count_nonzero(CDL_CV02 == 11))\n",
    "sensitivity02_y = np.divide(np.count_nonzero(CDL_CV02 == 0),denom_sens_02)\n",
    "denom_spec_02 = np.add(np.count_nonzero(CDL_CV02 == 10)- cdl111, np.count_nonzero(CDL_CV02 == -1))\n",
    "specificity02_x = np.divide(np.count_nonzero(CDL_CV02 == 10)- cdl111,denom_spec_02)\n",
    "sub_specificity02_x = np.subtract(1, specificity02_x)\n",
    "\n",
    "denom_sens_03 = np.add(np.count_nonzero(CDL_CV03 == 0), np.count_nonzero(CDL_CV03 == 11))\n",
    "sensitivity03_y = np.divide(np.count_nonzero(CDL_CV03 == 0),denom_sens_03)\n",
    "denom_spec_03 = np.add(np.count_nonzero(CDL_CV03 == 10)- cdl111, np.count_nonzero(CDL_CV03 == -1))\n",
    "specificity03_x = np.divide(np.count_nonzero(CDL_CV03 == 10)- cdl111,denom_spec_03)\n",
    "sub_specificity03_x = np.subtract(1, specificity03_x)\n",
    "\n",
    "denom_sens_04 = np.add(np.count_nonzero(CDL_CV04 == 0), np.count_nonzero(CDL_CV04 == 11))\n",
    "sensitivity04_y = np.divide(np.count_nonzero(CDL_CV04 == 0),denom_sens_04)\n",
    "denom_spec_04 = np.add(np.count_nonzero(CDL_CV04 == 10)- cdl111, np.count_nonzero(CDL_CV04 == -1))\n",
    "specificity04_x = np.divide(np.count_nonzero(CDL_CV04 == 10)- cdl111,denom_spec_04)\n",
    "sub_specificity04_x = np.subtract(1, specificity04_x)\n",
    "\n",
    "denom_sens_05 = np.add(np.count_nonzero(CDL_CV05 == 0), np.count_nonzero(CDL_CV05 == 11))\n",
    "sensitivity05_y = np.divide(np.count_nonzero(CDL_CV05 == 0),denom_sens_05)\n",
    "denom_spec_05 = np.add(np.count_nonzero(CDL_CV05 == 10)- cdl111, np.count_nonzero(CDL_CV05 == -1))\n",
    "specificity05_x = np.divide(np.count_nonzero(CDL_CV05 == 10)- cdl111,denom_spec_05)\n",
    "sub_specificity05_x = np.subtract(1, specificity05_x)\n",
    "\n",
    "denom_sens_06 = np.add(np.count_nonzero(CDL_CV06 == 0), np.count_nonzero(CDL_CV06 == 11))\n",
    "sensitivity06_y = np.divide(np.count_nonzero(CDL_CV06 == 0),denom_sens_06)\n",
    "denom_spec_06 = np.add(np.count_nonzero(CDL_CV06 == 10)- cdl111, np.count_nonzero(CDL_CV06 == -1))\n",
    "specificity06_x = np.divide(np.count_nonzero(CDL_CV06 == 10)- cdl111,denom_spec_06)\n",
    "sub_specificity06_x = np.subtract(1, specificity06_x)\n",
    "\n",
    "denom_sens_07 = np.add(np.count_nonzero(CDL_CV07 == 0), np.count_nonzero(CDL_CV07 == 11))\n",
    "sensitivity07_y = np.divide(np.count_nonzero(CDL_CV07 == 0),denom_sens_07)\n",
    "denom_spec_07 = np.add(np.count_nonzero(CDL_CV07 == 10)- cdl111, np.count_nonzero(CDL_CV07 == -1))\n",
    "specificity07_x = np.divide(np.count_nonzero(CDL_CV07 == 10)- cdl111,denom_spec_07)\n",
    "sub_specificity07_x = np.subtract(1, specificity07_x)\n",
    "\n",
    "denom_sens_08 = np.add(np.count_nonzero(CDL_CV08 == 0), np.count_nonzero(CDL_CV08 == 11))\n",
    "sensitivity08_y = np.divide(np.count_nonzero(CDL_CV08 == 0),denom_sens_08)\n",
    "denom_spec_08 = np.add(np.count_nonzero(CDL_CV08 == 10)- cdl111, np.count_nonzero(CDL_CV08 == -1))\n",
    "specificity08_x = np.divide(np.count_nonzero(CDL_CV08 == 10)- cdl111,denom_spec_08)\n",
    "sub_specificity08_x = np.subtract(1, specificity08_x)\n",
    "\n",
    "denom_sens_09 = np.add(np.count_nonzero(CDL_CV09 == 0), np.count_nonzero(CDL_CV09 == 11))\n",
    "sensitivity09_y = np.divide(np.count_nonzero(CDL_CV09 == 0),denom_sens_09)\n",
    "denom_spec_09 = np.add(np.count_nonzero(CDL_CV09 == 10)- cdl111, np.count_nonzero(CDL_CV09 == -1))\n",
    "specificity09_x = np.divide(np.count_nonzero(CDL_CV09 == 10)- cdl111,denom_spec_09)\n",
    "sub_specificity09_x = np.subtract(1, specificity09_x)\n",
    "\n",
    "#0.1 sensitivity and specificity \n",
    "denom_sens_1 = np.add(np.count_nonzero(CDL_CV1 == 0), np.count_nonzero(CDL_CV1 == 11))\n",
    "sensitivity1_y = np.divide(np.count_nonzero(CDL_CV1 == 0),denom_sens_1)\n",
    "denom_spec_1 = np.add(np.count_nonzero(CDL_CV1 == 10)- cdl111, np.count_nonzero(CDL_CV1 == -1))\n",
    "specificity1_x = np.divide(np.count_nonzero(CDL_CV1 == 10)- cdl111,denom_spec_1)\n",
    "sub_specificity1_x = np.subtract(1, specificity1_x)\n",
    "\n",
    "denom_sens_11 = np.add(np.count_nonzero(CDL_CV11 == 0), np.count_nonzero(CDL_CV11 == 11))\n",
    "sensitivity11_y = np.divide(np.count_nonzero(CDL_CV11 == 0),denom_sens_11)\n",
    "denom_spec_11 = np.add(np.count_nonzero(CDL_CV11 == 10)- cdl111, np.count_nonzero(CDL_CV11 == -1))\n",
    "specificity11_x = np.divide(np.count_nonzero(CDL_CV11 == 10)- cdl111,denom_spec_11)\n",
    "sub_specificity11_x = np.subtract(1, specificity11_x)\n",
    "\n",
    "denom_sens_12 = np.add(np.count_nonzero(CDL_CV12 == 0), np.count_nonzero(CDL_CV12 == 11))\n",
    "sensitivity12_y = np.divide(np.count_nonzero(CDL_CV12 == 0),denom_sens_12)\n",
    "denom_spec_12 = np.add(np.count_nonzero(CDL_CV12 == 10)- cdl111, np.count_nonzero(CDL_CV12 == -1))\n",
    "specificity12_x = np.divide(np.count_nonzero(CDL_CV12 == 10)- cdl111,denom_spec_12)\n",
    "sub_specificity12_x = np.subtract(1, specificity12_x)\n",
    "\n",
    "denom_sens_13 = np.add(np.count_nonzero(CDL_CV13 == 0), np.count_nonzero(CDL_CV13 == 11))\n",
    "sensitivity13_y = np.divide(np.count_nonzero(CDL_CV13 == 0),denom_sens_13)\n",
    "denom_spec_13 = np.add(np.count_nonzero(CDL_CV13 == 10)- cdl111, np.count_nonzero(CDL_CV13 == -1))\n",
    "specificity13_x = np.divide(np.count_nonzero(CDL_CV13 == 10)- cdl111,denom_spec_13)\n",
    "sub_specificity13_x = np.subtract(1, specificity13_x)\n",
    "\n",
    "denom_sens_14 = np.add(np.count_nonzero(CDL_CV14 == 0), np.count_nonzero(CDL_CV14 == 11))\n",
    "sensitivity14_y = np.divide(np.count_nonzero(CDL_CV14 == 0),denom_sens_14)\n",
    "denom_spec_14 = np.add(np.count_nonzero(CDL_CV14 == 10)- cdl111, np.count_nonzero(CDL_CV14 == -1))\n",
    "specificity14_x = np.divide(np.count_nonzero(CDL_CV14 == 10)- cdl111,denom_spec_14)\n",
    "sub_specificity14_x = np.subtract(1, specificity14_x)\n",
    "\n",
    "denom_sens_15 = np.add(np.count_nonzero(CDL_CV15 == 0), np.count_nonzero(CDL_CV15 == 11))\n",
    "sensitivity15_y = np.divide(np.count_nonzero(CDL_CV15 == 0),denom_sens_15)\n",
    "denom_spec_15 = np.add(np.count_nonzero(CDL_CV15 == 10)- cdl111, np.count_nonzero(CDL_CV15 == -1))\n",
    "specificity15_x = np.divide(np.count_nonzero(CDL_CV15 == 10)- cdl111,denom_spec_15)\n",
    "sub_specificity15_x = np.subtract(1, specificity15_x)\n",
    "\n",
    "denom_sens_16 = np.add(np.count_nonzero(CDL_CV16 == 0), np.count_nonzero(CDL_CV16 == 11))\n",
    "sensitivity16_y = np.divide(np.count_nonzero(CDL_CV16 == 0),denom_sens_16)\n",
    "denom_spec_16 = np.add(np.count_nonzero(CDL_CV16 == 10)- cdl111, np.count_nonzero(CDL_CV16 == -1))\n",
    "specificity16_x = np.divide(np.count_nonzero(CDL_CV16 == 10)- cdl111,denom_spec_16)\n",
    "sub_specificity16_x = np.subtract(1, specificity16_x)\n",
    "\n",
    "denom_sens_17 = np.add(np.count_nonzero(CDL_CV17 == 0), np.count_nonzero(CDL_CV17 == 11))\n",
    "sensitivity17_y = np.divide(np.count_nonzero(CDL_CV17 == 0),denom_sens_17)\n",
    "denom_spec_17 = np.add(np.count_nonzero(CDL_CV17 == 10)- cdl111, np.count_nonzero(CDL_CV17 == -1))\n",
    "specificity17_x = np.divide(np.count_nonzero(CDL_CV17 == 10)- cdl111,denom_spec_17)\n",
    "sub_specificity17_x = np.subtract(1, specificity17_x)\n",
    "\n",
    "denom_sens_18 = np.add(np.count_nonzero(CDL_CV18 == 0), np.count_nonzero(CDL_CV18 == 11))\n",
    "sensitivity18_y = np.divide(np.count_nonzero(CDL_CV18 == 0),denom_sens_18)\n",
    "denom_spec_18 = np.add(np.count_nonzero(CDL_CV18 == 10)- cdl111, np.count_nonzero(CDL_CV18 == -1))\n",
    "specificity18_x = np.divide(np.count_nonzero(CDL_CV18 == 10)- cdl111,denom_spec_18)\n",
    "sub_specificity18_x = np.subtract(1, specificity18_x)\n",
    "\n",
    "denom_sens_19 = np.add(np.count_nonzero(CDL_CV19 == 0), np.count_nonzero(CDL_CV19 == 11))\n",
    "sensitivity19_y = np.divide(np.count_nonzero(CDL_CV19 == 0),denom_sens_19)\n",
    "denom_spec_19 = np.add(np.count_nonzero(CDL_CV19 == 10)- cdl111, np.count_nonzero(CDL_CV19 == -1))\n",
    "specificity19_x = np.divide(np.count_nonzero(CDL_CV19 == 10)- cdl111,denom_spec_19)\n",
    "sub_specificity19_x = np.subtract(1, specificity19_x)\n",
    "\n",
    "#0.2 sensitivity and specificity \n",
    "denom_sens_2 = np.add(np.count_nonzero(CDL_CV2 == 0), np.count_nonzero(CDL_CV2 == 11))\n",
    "sensitivity2_y = np.divide(np.count_nonzero(CDL_CV2 == 0),denom_sens_2)\n",
    "denom_spec_2 = np.add(np.count_nonzero(CDL_CV2 == 10)- cdl111, np.count_nonzero(CDL_CV2 == -1))\n",
    "specificity2_x = np.divide(np.count_nonzero(CDL_CV2 == 10)- cdl111,denom_spec_2)\n",
    "sub_specificity2_x = np.subtract(1, specificity2_x)\n",
    "\n",
    "denom_sens_21 = np.add(np.count_nonzero(CDL_CV21 == 0), np.count_nonzero(CDL_CV21 == 11))\n",
    "sensitivity21_y = np.divide(np.count_nonzero(CDL_CV21 == 0),denom_sens_21)\n",
    "denom_spec_21 = np.add(np.count_nonzero(CDL_CV21 == 10)- cdl111, np.count_nonzero(CDL_CV21 == -1))\n",
    "specificity21_x = np.divide(np.count_nonzero(CDL_CV21 == 10)- cdl111,denom_spec_21)\n",
    "sub_specificity21_x = np.subtract(1, specificity21_x)\n",
    "\n",
    "denom_sens_22 = np.add(np.count_nonzero(CDL_CV22 == 0), np.count_nonzero(CDL_CV22 == 11))\n",
    "sensitivity22_y = np.divide(np.count_nonzero(CDL_CV22 == 0),denom_sens_22)\n",
    "denom_spec_22 = np.add(np.count_nonzero(CDL_CV22 == 10)- cdl111, np.count_nonzero(CDL_CV22 == -1))\n",
    "specificity22_x = np.divide(np.count_nonzero(CDL_CV22 == 10)- cdl111,denom_spec_22)\n",
    "sub_specificity22_x = np.subtract(1, specificity22_x)\n",
    "\n",
    "denom_sens_23 = np.add(np.count_nonzero(CDL_CV23 == 0), np.count_nonzero(CDL_CV23 == 11))\n",
    "sensitivity23_y = np.divide(np.count_nonzero(CDL_CV23 == 0),denom_sens_23)\n",
    "denom_spec_23 = np.add(np.count_nonzero(CDL_CV23 == 10)- cdl111, np.count_nonzero(CDL_CV23 == -1))\n",
    "specificity23_x = np.divide(np.count_nonzero(CDL_CV23 == 10)- cdl111,denom_spec_23)\n",
    "sub_specificity23_x = np.subtract(1, specificity23_x)\n",
    "\n",
    "denom_sens_24 = np.add(np.count_nonzero(CDL_CV24 == 0), np.count_nonzero(CDL_CV24 == 11))\n",
    "sensitivity24_y = np.divide(np.count_nonzero(CDL_CV24 == 0),denom_sens_24)\n",
    "denom_spec_24 = np.add(np.count_nonzero(CDL_CV24 == 10)- cdl111, np.count_nonzero(CDL_CV24 == -1))\n",
    "specificity24_x = np.divide(np.count_nonzero(CDL_CV24 == 10)- cdl111,denom_spec_24)\n",
    "sub_specificity24_x = np.subtract(1, specificity24_x)\n",
    "\n",
    "denom_sens_25 = np.add(np.count_nonzero(CDL_CV25 == 0), np.count_nonzero(CDL_CV25 == 11))\n",
    "sensitivity25_y = np.divide(np.count_nonzero(CDL_CV25 == 0),denom_sens_25)\n",
    "denom_spec_25 = np.add(np.count_nonzero(CDL_CV25 == 10)- cdl111, np.count_nonzero(CDL_CV25 == -1))\n",
    "specificity25_x = np.divide(np.count_nonzero(CDL_CV25 == 10)- cdl111,denom_spec_25)\n",
    "sub_specificity25_x = np.subtract(1, specificity25_x)\n",
    "\n",
    "denom_sens_26 = np.add(np.count_nonzero(CDL_CV26 == 0), np.count_nonzero(CDL_CV26 == 11))\n",
    "sensitivity26_y = np.divide(np.count_nonzero(CDL_CV26 == 0),denom_sens_26)\n",
    "denom_spec_26 = np.add(np.count_nonzero(CDL_CV26 == 10)- cdl111, np.count_nonzero(CDL_CV26 == -1))\n",
    "specificity26_x = np.divide(np.count_nonzero(CDL_CV26 == 10)- cdl111,denom_spec_26)\n",
    "sub_specificity26_x = np.subtract(1, specificity26_x)\n",
    "\n",
    "denom_sens_27 = np.add(np.count_nonzero(CDL_CV27 == 0), np.count_nonzero(CDL_CV27 == 11))\n",
    "sensitivity27_y = np.divide(np.count_nonzero(CDL_CV27 == 0),denom_sens_27)\n",
    "denom_spec_27 = np.add(np.count_nonzero(CDL_CV27 == 10)- cdl111, np.count_nonzero(CDL_CV27 == -1))\n",
    "specificity27_x = np.divide(np.count_nonzero(CDL_CV27 == 10)- cdl111,denom_spec_27)\n",
    "sub_specificity27_x = np.subtract(1, specificity27_x)\n",
    "\n",
    "denom_sens_28 = np.add(np.count_nonzero(CDL_CV28 == 0), np.count_nonzero(CDL_CV28 == 11))\n",
    "sensitivity28_y = np.divide(np.count_nonzero(CDL_CV28 == 0),denom_sens_28)\n",
    "denom_spec_28 = np.add(np.count_nonzero(CDL_CV28 == 10)- cdl111, np.count_nonzero(CDL_CV28 == -1))\n",
    "specificity28_x = np.divide(np.count_nonzero(CDL_CV28 == 10)- cdl111,denom_spec_28)\n",
    "sub_specificity28_x = np.subtract(1, specificity28_x)\n",
    "\n",
    "denom_sens_29 = np.add(np.count_nonzero(CDL_CV29 == 0), np.count_nonzero(CDL_CV29 == 11))\n",
    "sensitivity29_y = np.divide(np.count_nonzero(CDL_CV29 == 0),denom_sens_29)\n",
    "denom_spec_29 = np.add(np.count_nonzero(CDL_CV29 == 10)- cdl111, np.count_nonzero(CDL_CV29 == -1))\n",
    "specificity29_x = np.divide(np.count_nonzero(CDL_CV29 == 10)- cdl111,denom_spec_29)\n",
    "sub_specificity29_x = np.subtract(1, specificity29_x)\n",
    "\n",
    "#0.3 sensitivity and specificity \n",
    "denom_sens_3 = np.add(np.count_nonzero(CDL_CV3 == 0), np.count_nonzero(CDL_CV3 == 11))\n",
    "sensitivity3_y = np.divide(np.count_nonzero(CDL_CV3 == 0),denom_sens_3)\n",
    "denom_spec_3 = np.add(np.count_nonzero(CDL_CV3 == 10)- cdl111, np.count_nonzero(CDL_CV3 == -1))\n",
    "specificity3_x = np.divide(np.count_nonzero(CDL_CV3 == 10)- cdl111,denom_spec_3)\n",
    "sub_specificity3_x = np.subtract(1, specificity3_x)\n",
    "\n",
    "denom_sens_31 = np.add(np.count_nonzero(CDL_CV31 == 0), np.count_nonzero(CDL_CV31 == 11))\n",
    "sensitivity31_y = np.divide(np.count_nonzero(CDL_CV31 == 0),denom_sens_31)\n",
    "denom_spec_31 = np.add(np.count_nonzero(CDL_CV31 == 10)- cdl111, np.count_nonzero(CDL_CV31 == -1))\n",
    "specificity31_x = np.divide(np.count_nonzero(CDL_CV31 == 10)- cdl111,denom_spec_31)\n",
    "sub_specificity31_x = np.subtract(1, specificity31_x)\n",
    "\n",
    "denom_sens_32 = np.add(np.count_nonzero(CDL_CV32 == 0), np.count_nonzero(CDL_CV32 == 11))\n",
    "sensitivity32_y = np.divide(np.count_nonzero(CDL_CV32 == 0),denom_sens_32)\n",
    "denom_spec_32 = np.add(np.count_nonzero(CDL_CV32 == 10)- cdl111, np.count_nonzero(CDL_CV32 == -1))\n",
    "specificity32_x = np.divide(np.count_nonzero(CDL_CV32 == 10)- cdl111,denom_spec_32)\n",
    "sub_specificity32_x = np.subtract(1, specificity32_x)\n",
    "\n",
    "denom_sens_33 = np.add(np.count_nonzero(CDL_CV33 == 0), np.count_nonzero(CDL_CV33 == 11))\n",
    "sensitivity33_y = np.divide(np.count_nonzero(CDL_CV33 == 0),denom_sens_33)\n",
    "denom_spec_33 = np.add(np.count_nonzero(CDL_CV33 == 10)- cdl111, np.count_nonzero(CDL_CV33 == -1))\n",
    "specificity33_x = np.divide(np.count_nonzero(CDL_CV33 == 10)- cdl111,denom_spec_33)\n",
    "sub_specificity33_x = np.subtract(1, specificity33_x)\n",
    "\n",
    "denom_sens_34 = np.add(np.count_nonzero(CDL_CV34 == 0), np.count_nonzero(CDL_CV34 == 11))\n",
    "sensitivity34_y = np.divide(np.count_nonzero(CDL_CV34 == 0),denom_sens_34)\n",
    "denom_spec_34 = np.add(np.count_nonzero(CDL_CV34 == 10)- cdl111, np.count_nonzero(CDL_CV34 == -1))\n",
    "specificity34_x = np.divide(np.count_nonzero(CDL_CV34 == 10)- cdl111,denom_spec_34)\n",
    "sub_specificity34_x = np.subtract(1, specificity34_x)\n",
    "\n",
    "denom_sens_35 = np.add(np.count_nonzero(CDL_CV35 == 0), np.count_nonzero(CDL_CV35 == 11))\n",
    "sensitivity35_y = np.divide(np.count_nonzero(CDL_CV35 == 0),denom_sens_35)\n",
    "denom_spec_35 = np.add(np.count_nonzero(CDL_CV35 == 10)- cdl111, np.count_nonzero(CDL_CV35 == -1))\n",
    "specificity35_x = np.divide(np.count_nonzero(CDL_CV35 == 10)- cdl111,denom_spec_35)\n",
    "sub_specificity35_x = np.subtract(1, specificity35_x)\n",
    "\n",
    "denom_sens_36 = np.add(np.count_nonzero(CDL_CV36 == 0), np.count_nonzero(CDL_CV36 == 11))\n",
    "sensitivity36_y = np.divide(np.count_nonzero(CDL_CV36 == 0),denom_sens_36)\n",
    "denom_spec_36 = np.add(np.count_nonzero(CDL_CV36 == 10)- cdl111, np.count_nonzero(CDL_CV36 == -1))\n",
    "specificity36_x = np.divide(np.count_nonzero(CDL_CV36 == 10)- cdl111,denom_spec_36)\n",
    "sub_specificity36_x = np.subtract(1, specificity36_x)\n",
    "\n",
    "denom_sens_37 = np.add(np.count_nonzero(CDL_CV37 == 0), np.count_nonzero(CDL_CV37 == 11))\n",
    "sensitivity37_y = np.divide(np.count_nonzero(CDL_CV37 == 0),denom_sens_37)\n",
    "denom_spec_37 = np.add(np.count_nonzero(CDL_CV37 == 10)- cdl111, np.count_nonzero(CDL_CV37 == -1))\n",
    "specificity37_x = np.divide(np.count_nonzero(CDL_CV37 == 10)- cdl111,denom_spec_37)\n",
    "sub_specificity37_x = np.subtract(1, specificity37_x)\n",
    "\n",
    "denom_sens_38 = np.add(np.count_nonzero(CDL_CV38 == 0), np.count_nonzero(CDL_CV38 == 11))\n",
    "sensitivity38_y = np.divide(np.count_nonzero(CDL_CV38 == 0),denom_sens_38)\n",
    "denom_spec_38 = np.add(np.count_nonzero(CDL_CV38 == 10)- cdl111, np.count_nonzero(CDL_CV38 == -1))\n",
    "specificity38_x = np.divide(np.count_nonzero(CDL_CV38 == 10)- cdl111,denom_spec_38)\n",
    "sub_specificity38_x = np.subtract(1, specificity38_x)\n",
    "\n",
    "denom_sens_39 = np.add(np.count_nonzero(CDL_CV39 == 0), np.count_nonzero(CDL_CV39 == 11))\n",
    "sensitivity39_y = np.divide(np.count_nonzero(CDL_CV39 == 0),denom_sens_39)\n",
    "denom_spec_39 = np.add(np.count_nonzero(CDL_CV39 == 10)- cdl111, np.count_nonzero(CDL_CV39 == -1))\n",
    "specificity39_x = np.divide(np.count_nonzero(CDL_CV39 == 10)- cdl111,denom_spec_39)\n",
    "sub_specificity39_x = np.subtract(1, specificity39_x)\n",
    "\n",
    "#0.4 sensitivity and specificity \n",
    "denom_sens_4 = np.add(np.count_nonzero(CDL_CV4 == 0), np.count_nonzero(CDL_CV4 == 11))\n",
    "sensitivity4_y = np.divide(np.count_nonzero(CDL_CV4 == 0),denom_sens_4)\n",
    "denom_spec_4 = np.add(np.count_nonzero(CDL_CV4 == 10)- cdl111, np.count_nonzero(CDL_CV4 == -1))\n",
    "specificity4_x = np.divide(np.count_nonzero(CDL_CV4 == 10)- cdl111,denom_spec_4)\n",
    "sub_specificity4_x = np.subtract(1, specificity4_x)\n",
    "\n",
    "denom_sens_41 = np.add(np.count_nonzero(CDL_CV41 == 0), np.count_nonzero(CDL_CV41 == 11))\n",
    "sensitivity41_y = np.divide(np.count_nonzero(CDL_CV41 == 0),denom_sens_41)\n",
    "denom_spec_41 = np.add(np.count_nonzero(CDL_CV41 == 10)- cdl111, np.count_nonzero(CDL_CV41 == -1))\n",
    "specificity41_x = np.divide(np.count_nonzero(CDL_CV41 == 10)- cdl111,denom_spec_41)\n",
    "sub_specificity41_x = np.subtract(1, specificity41_x)\n",
    "\n",
    "denom_sens_42 = np.add(np.count_nonzero(CDL_CV42 == 0), np.count_nonzero(CDL_CV42 == 11))\n",
    "sensitivity42_y = np.divide(np.count_nonzero(CDL_CV42 == 0),denom_sens_42)\n",
    "denom_spec_42 = np.add(np.count_nonzero(CDL_CV42 == 10)- cdl111, np.count_nonzero(CDL_CV42 == -1))\n",
    "specificity42_x = np.divide(np.count_nonzero(CDL_CV42 == 10)- cdl111,denom_spec_42)\n",
    "sub_specificity42_x = np.subtract(1, specificity42_x)\n",
    "\n",
    "denom_sens_43 = np.add(np.count_nonzero(CDL_CV43 == 0), np.count_nonzero(CDL_CV43 == 11))\n",
    "sensitivity43_y = np.divide(np.count_nonzero(CDL_CV43 == 0),denom_sens_43)\n",
    "denom_spec_43 = np.add(np.count_nonzero(CDL_CV43 == 10)- cdl111, np.count_nonzero(CDL_CV43 == -1))\n",
    "specificity43_x = np.divide(np.count_nonzero(CDL_CV43 == 10)- cdl111,denom_spec_43)\n",
    "sub_specificity43_x = np.subtract(1, specificity43_x)\n",
    "\n",
    "denom_sens_44 = np.add(np.count_nonzero(CDL_CV44 == 0), np.count_nonzero(CDL_CV44 == 11))\n",
    "sensitivity44_y = np.divide(np.count_nonzero(CDL_CV44 == 0),denom_sens_44)\n",
    "denom_spec_44 = np.add(np.count_nonzero(CDL_CV44 == 10)- cdl111, np.count_nonzero(CDL_CV44 == -1))\n",
    "specificity44_x = np.divide(np.count_nonzero(CDL_CV44 == 10)- cdl111,denom_spec_44)\n",
    "sub_specificity44_x = np.subtract(1, specificity44_x)\n",
    "\n",
    "denom_sens_45 = np.add(np.count_nonzero(CDL_CV45 == 0), np.count_nonzero(CDL_CV45 == 11))\n",
    "sensitivity45_y = np.divide(np.count_nonzero(CDL_CV45 == 0),denom_sens_45)\n",
    "denom_spec_45 = np.add(np.count_nonzero(CDL_CV45 == 10)- cdl111, np.count_nonzero(CDL_CV45 == -1))\n",
    "specificity45_x = np.divide(np.count_nonzero(CDL_CV45 == 10)- cdl111,denom_spec_45)\n",
    "sub_specificity45_x = np.subtract(1, specificity45_x)\n",
    "\n",
    "denom_sens_46 = np.add(np.count_nonzero(CDL_CV46 == 0), np.count_nonzero(CDL_CV46 == 11))\n",
    "sensitivity46_y = np.divide(np.count_nonzero(CDL_CV46 == 0),denom_sens_46)\n",
    "denom_spec_46 = np.add(np.count_nonzero(CDL_CV46 == 10)- cdl111, np.count_nonzero(CDL_CV46 == -1))\n",
    "specificity46_x = np.divide(np.count_nonzero(CDL_CV46 == 10)- cdl111,denom_spec_46)\n",
    "sub_specificity46_x = np.subtract(1, specificity46_x)\n",
    "\n",
    "denom_sens_47 = np.add(np.count_nonzero(CDL_CV47 == 0), np.count_nonzero(CDL_CV47 == 11))\n",
    "sensitivity47_y = np.divide(np.count_nonzero(CDL_CV47 == 0),denom_sens_47)\n",
    "denom_spec_47 = np.add(np.count_nonzero(CDL_CV47 == 10)- cdl111, np.count_nonzero(CDL_CV47 == -1))\n",
    "specificity47_x = np.divide(np.count_nonzero(CDL_CV47 == 10)- cdl111,denom_spec_47)\n",
    "sub_specificity47_x = np.subtract(1, specificity47_x)\n",
    "\n",
    "denom_sens_48 = np.add(np.count_nonzero(CDL_CV48 == 0), np.count_nonzero(CDL_CV48 == 11))\n",
    "sensitivity48_y = np.divide(np.count_nonzero(CDL_CV48 == 0),denom_sens_48)\n",
    "denom_spec_48 = np.add(np.count_nonzero(CDL_CV48 == 10)- cdl111, np.count_nonzero(CDL_CV48 == -1))\n",
    "specificity48_x = np.divide(np.count_nonzero(CDL_CV48 == 10)- cdl111,denom_spec_48)\n",
    "sub_specificity48_x = np.subtract(1, specificity48_x)\n",
    "\n",
    "denom_sens_49 = np.add(np.count_nonzero(CDL_CV49 == 0), np.count_nonzero(CDL_CV49 == 11))\n",
    "sensitivity49_y = np.divide(np.count_nonzero(CDL_CV49 == 0),denom_sens_49)\n",
    "denom_spec_49 = np.add(np.count_nonzero(CDL_CV49 == 10)- cdl111, np.count_nonzero(CDL_CV49 == -1))\n",
    "specificity49_x = np.divide(np.count_nonzero(CDL_CV49 == 10)- cdl111,denom_spec_49)\n",
    "sub_specificity49_x = np.subtract(1, specificity49_x)\n",
    "\n",
    "#0.5 sensitivity and specificity \n",
    "denom_sens_5 = np.add(np.count_nonzero(CDL_CV5 == 0), np.count_nonzero(CDL_CV5 == 11))\n",
    "sensitivity5_y = np.divide(np.count_nonzero(CDL_CV5 == 0),denom_sens_5)\n",
    "denom_spec_5 = np.add(np.count_nonzero(CDL_CV5 == 10)- cdl111, np.count_nonzero(CDL_CV5 == -1))\n",
    "specificity5_x = np.divide(np.count_nonzero(CDL_CV5 == 10)- cdl111,denom_spec_5)\n",
    "sub_specificity5_x = np.subtract(1, specificity5_x)\n",
    "\n",
    "denom_sens_51 = np.add(np.count_nonzero(CDL_CV51 == 0), np.count_nonzero(CDL_CV51 == 11))\n",
    "sensitivity51_y = np.divide(np.count_nonzero(CDL_CV51 == 0),denom_sens_51)\n",
    "denom_spec_51 = np.add(np.count_nonzero(CDL_CV51 == 10)- cdl111, np.count_nonzero(CDL_CV51 == -1))\n",
    "specificity51_x = np.divide(np.count_nonzero(CDL_CV51 == 10)- cdl111,denom_spec_51)\n",
    "sub_specificity51_x = np.subtract(1, specificity51_x)\n",
    "\n",
    "denom_sens_52 = np.add(np.count_nonzero(CDL_CV52 == 0), np.count_nonzero(CDL_CV52 == 11))\n",
    "sensitivity52_y = np.divide(np.count_nonzero(CDL_CV52 == 0),denom_sens_52)\n",
    "denom_spec_52 = np.add(np.count_nonzero(CDL_CV52 == 10)- cdl111, np.count_nonzero(CDL_CV52 == -1))\n",
    "specificity52_x = np.divide(np.count_nonzero(CDL_CV52 == 10)- cdl111,denom_spec_52)\n",
    "sub_specificity52_x = np.subtract(1, specificity52_x)\n",
    "\n",
    "denom_sens_53 = np.add(np.count_nonzero(CDL_CV53 == 0), np.count_nonzero(CDL_CV53 == 11))\n",
    "sensitivity53_y = np.divide(np.count_nonzero(CDL_CV53 == 0),denom_sens_53)\n",
    "denom_spec_53 = np.add(np.count_nonzero(CDL_CV53 == 10)- cdl111, np.count_nonzero(CDL_CV53 == -1))\n",
    "specificity53_x = np.divide(np.count_nonzero(CDL_CV53 == 10)- cdl111,denom_spec_53)\n",
    "sub_specificity53_x = np.subtract(1, specificity53_x)\n",
    "\n",
    "denom_sens_54 = np.add(np.count_nonzero(CDL_CV54 == 0), np.count_nonzero(CDL_CV54 == 11))\n",
    "sensitivity54_y = np.divide(np.count_nonzero(CDL_CV54 == 0),denom_sens_54)\n",
    "denom_spec_54 = np.add(np.count_nonzero(CDL_CV54 == 10)- cdl111, np.count_nonzero(CDL_CV54 == -1))\n",
    "specificity54_x = np.divide(np.count_nonzero(CDL_CV54 == 10)- cdl111,denom_spec_54)\n",
    "sub_specificity54_x = np.subtract(1, specificity54_x)\n",
    "\n",
    "denom_sens_55 = np.add(np.count_nonzero(CDL_CV55 == 0), np.count_nonzero(CDL_CV55 == 11))\n",
    "sensitivity55_y = np.divide(np.count_nonzero(CDL_CV55 == 0),denom_sens_55)\n",
    "denom_spec_55 = np.add(np.count_nonzero(CDL_CV55 == 10)- cdl111, np.count_nonzero(CDL_CV55 == -1))\n",
    "specificity55_x = np.divide(np.count_nonzero(CDL_CV55 == 10)- cdl111,denom_spec_55)\n",
    "sub_specificity55_x = np.subtract(1, specificity55_x)\n",
    "\n",
    "denom_sens_56 = np.add(np.count_nonzero(CDL_CV56 == 0), np.count_nonzero(CDL_CV56 == 11))\n",
    "sensitivity56_y = np.divide(np.count_nonzero(CDL_CV56 == 0),denom_sens_56)\n",
    "denom_spec_56 = np.add(np.count_nonzero(CDL_CV56 == 10)- cdl111, np.count_nonzero(CDL_CV56 == -1))\n",
    "specificity56_x = np.divide(np.count_nonzero(CDL_CV56 == 10)- cdl111,denom_spec_56)\n",
    "sub_specificity56_x = np.subtract(1, specificity56_x)\n",
    "\n",
    "denom_sens_57 = np.add(np.count_nonzero(CDL_CV57 == 0), np.count_nonzero(CDL_CV57 == 11))\n",
    "sensitivity57_y = np.divide(np.count_nonzero(CDL_CV57 == 0),denom_sens_57)\n",
    "denom_spec_57 = np.add(np.count_nonzero(CDL_CV57 == 10)- cdl111, np.count_nonzero(CDL_CV57 == -1))\n",
    "specificity57_x = np.divide(np.count_nonzero(CDL_CV57 == 10)- cdl111,denom_spec_57)\n",
    "sub_specificity57_x = np.subtract(1, specificity57_x)\n",
    "\n",
    "denom_sens_58 = np.add(np.count_nonzero(CDL_CV58 == 0), np.count_nonzero(CDL_CV58 == 11))\n",
    "sensitivity58_y = np.divide(np.count_nonzero(CDL_CV58 == 0),denom_sens_58)\n",
    "denom_spec_58 = np.add(np.count_nonzero(CDL_CV58 == 10)- cdl111, np.count_nonzero(CDL_CV58 == -1))\n",
    "specificity58_x = np.divide(np.count_nonzero(CDL_CV58 == 10)- cdl111,denom_spec_58)\n",
    "sub_specificity58_x = np.subtract(1, specificity58_x)\n",
    "\n",
    "denom_sens_59 = np.add(np.count_nonzero(CDL_CV59 == 0), np.count_nonzero(CDL_CV59 == 11))\n",
    "sensitivity59_y = np.divide(np.count_nonzero(CDL_CV59 == 0),denom_sens_59)\n",
    "denom_spec_59 = np.add(np.count_nonzero(CDL_CV59 == 10)- cdl111, np.count_nonzero(CDL_CV59 == -1))\n",
    "specificity59_x = np.divide(np.count_nonzero(CDL_CV59 == 10)- cdl111,denom_spec_59)\n",
    "sub_specificity59_x = np.subtract(1, specificity59_x)\n",
    "\n",
    "#0.6 sensitivity and specificity \n",
    "denom_sens_6 = np.add(np.count_nonzero(CDL_CV6 == 0), np.count_nonzero(CDL_CV6 == 11))\n",
    "sensitivity6_y = np.divide(np.count_nonzero(CDL_CV6 == 0),denom_sens_6)\n",
    "denom_spec_6 = np.add(np.count_nonzero(CDL_CV6 == 10)- cdl111, np.count_nonzero(CDL_CV6 == -1))\n",
    "specificity6_x = np.divide(np.count_nonzero(CDL_CV6 == 10)- cdl111,denom_spec_6)\n",
    "sub_specificity6_x = np.subtract(1, specificity6_x)\n",
    "\n",
    "denom_sens_61 = np.add(np.count_nonzero(CDL_CV61 == 0), np.count_nonzero(CDL_CV61 == 11))\n",
    "sensitivity61_y = np.divide(np.count_nonzero(CDL_CV61 == 0),denom_sens_61)\n",
    "denom_spec_61 = np.add(np.count_nonzero(CDL_CV61 == 10)- cdl111, np.count_nonzero(CDL_CV61 == -1))\n",
    "specificity61_x = np.divide(np.count_nonzero(CDL_CV61 == 10)- cdl111,denom_spec_61)\n",
    "sub_specificity61_x = np.subtract(1, specificity61_x)\n",
    "\n",
    "denom_sens_62 = np.add(np.count_nonzero(CDL_CV62 == 0), np.count_nonzero(CDL_CV62 == 11))\n",
    "sensitivity62_y = np.divide(np.count_nonzero(CDL_CV62 == 0),denom_sens_62)\n",
    "denom_spec_62 = np.add(np.count_nonzero(CDL_CV62 == 10)- cdl111, np.count_nonzero(CDL_CV62 == -1))\n",
    "specificity62_x = np.divide(np.count_nonzero(CDL_CV62 == 10)- cdl111,denom_spec_62)\n",
    "sub_specificity62_x = np.subtract(1, specificity62_x)\n",
    "\n",
    "denom_sens_63 = np.add(np.count_nonzero(CDL_CV63 == 0), np.count_nonzero(CDL_CV63 == 11))\n",
    "sensitivity63_y = np.divide(np.count_nonzero(CDL_CV63 == 0),denom_sens_63)\n",
    "denom_spec_63 = np.add(np.count_nonzero(CDL_CV63 == 10)- cdl111, np.count_nonzero(CDL_CV63 == -1))\n",
    "specificity63_x = np.divide(np.count_nonzero(CDL_CV63 == 10)- cdl111,denom_spec_63)\n",
    "sub_specificity63_x = np.subtract(1, specificity63_x)\n",
    "\n",
    "denom_sens_64 = np.add(np.count_nonzero(CDL_CV64 == 0), np.count_nonzero(CDL_CV64 == 11))\n",
    "sensitivity64_y = np.divide(np.count_nonzero(CDL_CV64 == 0),denom_sens_64)\n",
    "denom_spec_64 = np.add(np.count_nonzero(CDL_CV64 == 10)- cdl111, np.count_nonzero(CDL_CV64 == -1))\n",
    "specificity64_x = np.divide(np.count_nonzero(CDL_CV64 == 10)- cdl111,denom_spec_64)\n",
    "sub_specificity64_x = np.subtract(1, specificity64_x)\n",
    "\n",
    "denom_sens_65 = np.add(np.count_nonzero(CDL_CV65 == 0), np.count_nonzero(CDL_CV65 == 11))\n",
    "sensitivity65_y = np.divide(np.count_nonzero(CDL_CV65 == 0),denom_sens_65)\n",
    "denom_spec_65 = np.add(np.count_nonzero(CDL_CV65 == 10)- cdl111, np.count_nonzero(CDL_CV65 == -1))\n",
    "specificity65_x = np.divide(np.count_nonzero(CDL_CV65 == 10)- cdl111,denom_spec_65)\n",
    "sub_specificity65_x = np.subtract(1, specificity65_x)\n",
    "\n",
    "denom_sens_66 = np.add(np.count_nonzero(CDL_CV66 == 0), np.count_nonzero(CDL_CV66 == 11))\n",
    "sensitivity66_y = np.divide(np.count_nonzero(CDL_CV66 == 0),denom_sens_66)\n",
    "denom_spec_66 = np.add(np.count_nonzero(CDL_CV66 == 10)- cdl111, np.count_nonzero(CDL_CV66 == -1))\n",
    "specificity66_x = np.divide(np.count_nonzero(CDL_CV66 == 10)- cdl111,denom_spec_66)\n",
    "sub_specificity66_x = np.subtract(1, specificity66_x)\n",
    "\n",
    "denom_sens_67 = np.add(np.count_nonzero(CDL_CV67 == 0), np.count_nonzero(CDL_CV67 == 11))\n",
    "sensitivity67_y = np.divide(np.count_nonzero(CDL_CV67 == 0),denom_sens_67)\n",
    "denom_spec_67 = np.add(np.count_nonzero(CDL_CV67 == 10)- cdl111, np.count_nonzero(CDL_CV67 == -1))\n",
    "specificity67_x = np.divide(np.count_nonzero(CDL_CV67 == 10)- cdl111,denom_spec_67)\n",
    "sub_specificity67_x = np.subtract(1, specificity67_x)\n",
    "\n",
    "denom_sens_68 = np.add(np.count_nonzero(CDL_CV68 == 0), np.count_nonzero(CDL_CV68 == 11))\n",
    "sensitivity68_y = np.divide(np.count_nonzero(CDL_CV68 == 0),denom_sens_68)\n",
    "denom_spec_68 = np.add(np.count_nonzero(CDL_CV68 == 10)- cdl111, np.count_nonzero(CDL_CV68 == -1))\n",
    "specificity68_x = np.divide(np.count_nonzero(CDL_CV68 == 10)- cdl111,denom_spec_68)\n",
    "sub_specificity68_x = np.subtract(1, specificity68_x)\n",
    "\n",
    "denom_sens_69 = np.add(np.count_nonzero(CDL_CV69 == 0), np.count_nonzero(CDL_CV69 == 11))\n",
    "sensitivity69_y = np.divide(np.count_nonzero(CDL_CV69 == 0),denom_sens_69)\n",
    "denom_spec_69 = np.add(np.count_nonzero(CDL_CV69 == 10)- cdl111, np.count_nonzero(CDL_CV69 == -1))\n",
    "specificity69_x = np.divide(np.count_nonzero(CDL_CV69 == 10)- cdl111,denom_spec_69)\n",
    "sub_specificity69_x = np.subtract(1, specificity69_x)\n",
    "\n",
    "#0.7 sensitivity and specificity \n",
    "denom_sens_7 = np.add(np.count_nonzero(CDL_CV7 == 0), np.count_nonzero(CDL_CV7 == 11))\n",
    "sensitivity7_y = np.divide(np.count_nonzero(CDL_CV7 == 0),denom_sens_7)\n",
    "denom_spec_7 = np.add(np.count_nonzero(CDL_CV7 == 10)- cdl111, np.count_nonzero(CDL_CV7 == -1))\n",
    "specificity7_x = np.divide(np.count_nonzero(CDL_CV7 == 10)- cdl111,denom_spec_7)\n",
    "sub_specificity7_x = np.subtract(1, specificity7_x)\n",
    "\n",
    "denom_sens_71 = np.add(np.count_nonzero(CDL_CV71 == 0), np.count_nonzero(CDL_CV71 == 11))\n",
    "sensitivity71_y = np.divide(np.count_nonzero(CDL_CV71 == 0),denom_sens_71)\n",
    "denom_spec_71 = np.add(np.count_nonzero(CDL_CV71 == 10)- cdl111, np.count_nonzero(CDL_CV71 == -1))\n",
    "specificity71_x = np.divide(np.count_nonzero(CDL_CV71 == 10)- cdl111,denom_spec_71)\n",
    "sub_specificity71_x = np.subtract(1, specificity71_x)\n",
    "\n",
    "denom_sens_72 = np.add(np.count_nonzero(CDL_CV72 == 0), np.count_nonzero(CDL_CV72 == 11))\n",
    "sensitivity72_y = np.divide(np.count_nonzero(CDL_CV72 == 0),denom_sens_72)\n",
    "denom_spec_72 = np.add(np.count_nonzero(CDL_CV72 == 10)- cdl111, np.count_nonzero(CDL_CV72 == -1))\n",
    "specificity72_x = np.divide(np.count_nonzero(CDL_CV72 == 10)- cdl111,denom_spec_72)\n",
    "sub_specificity72_x = np.subtract(1, specificity72_x)\n",
    "\n",
    "denom_sens_73 = np.add(np.count_nonzero(CDL_CV73 == 0), np.count_nonzero(CDL_CV73 == 11))\n",
    "sensitivity73_y = np.divide(np.count_nonzero(CDL_CV73 == 0),denom_sens_73)\n",
    "denom_spec_73 = np.add(np.count_nonzero(CDL_CV73 == 10)- cdl111, np.count_nonzero(CDL_CV73 == -1))\n",
    "specificity73_x = np.divide(np.count_nonzero(CDL_CV73 == 10)- cdl111,denom_spec_73)\n",
    "sub_specificity73_x = np.subtract(1, specificity73_x)\n",
    "\n",
    "denom_sens_74 = np.add(np.count_nonzero(CDL_CV74 == 0), np.count_nonzero(CDL_CV74 == 11))\n",
    "sensitivity74_y = np.divide(np.count_nonzero(CDL_CV74 == 0),denom_sens_74)\n",
    "denom_spec_74 = np.add(np.count_nonzero(CDL_CV74 == 10)- cdl111, np.count_nonzero(CDL_CV74 == -1))\n",
    "specificity74_x = np.divide(np.count_nonzero(CDL_CV74 == 10)- cdl111,denom_spec_74)\n",
    "sub_specificity74_x = np.subtract(1, specificity74_x)\n",
    "\n",
    "denom_sens_75 = np.add(np.count_nonzero(CDL_CV75 == 0), np.count_nonzero(CDL_CV75 == 11))\n",
    "sensitivity75_y = np.divide(np.count_nonzero(CDL_CV75 == 0),denom_sens_75)\n",
    "denom_spec_75 = np.add(np.count_nonzero(CDL_CV75 == 10)- cdl111, np.count_nonzero(CDL_CV75 == -1))\n",
    "specificity75_x = np.divide(np.count_nonzero(CDL_CV75 == 10)- cdl111,denom_spec_75)\n",
    "sub_specificity75_x = np.subtract(1, specificity75_x)\n",
    "\n",
    "denom_sens_76 = np.add(np.count_nonzero(CDL_CV76 == 0), np.count_nonzero(CDL_CV76 == 11))\n",
    "sensitivity76_y = np.divide(np.count_nonzero(CDL_CV76 == 0),denom_sens_76)\n",
    "denom_spec_76 = np.add(np.count_nonzero(CDL_CV76 == 10)- cdl111, np.count_nonzero(CDL_CV76 == -1))\n",
    "specificity76_x = np.divide(np.count_nonzero(CDL_CV76 == 10)- cdl111,denom_spec_76)\n",
    "sub_specificity76_x = np.subtract(1, specificity76_x)\n",
    "\n",
    "denom_sens_77 = np.add(np.count_nonzero(CDL_CV77 == 0), np.count_nonzero(CDL_CV77 == 11))\n",
    "sensitivity77_y = np.divide(np.count_nonzero(CDL_CV77 == 0),denom_sens_77)\n",
    "denom_spec_77 = np.add(np.count_nonzero(CDL_CV77 == 10)- cdl111, np.count_nonzero(CDL_CV77 == -1))\n",
    "specificity77_x = np.divide(np.count_nonzero(CDL_CV77 == 10)- cdl111,denom_spec_77)\n",
    "sub_specificity77_x = np.subtract(1, specificity77_x)\n",
    "\n",
    "denom_sens_78 = np.add(np.count_nonzero(CDL_CV78 == 0), np.count_nonzero(CDL_CV78 == 11))\n",
    "sensitivity78_y = np.divide(np.count_nonzero(CDL_CV78 == 0),denom_sens_78)\n",
    "denom_spec_78 = np.add(np.count_nonzero(CDL_CV78 == 10)- cdl111, np.count_nonzero(CDL_CV78 == -1))\n",
    "specificity78_x = np.divide(np.count_nonzero(CDL_CV78 == 10)- cdl111,denom_spec_78)\n",
    "sub_specificity78_x = np.subtract(1, specificity78_x)\n",
    "\n",
    "denom_sens_79 = np.add(np.count_nonzero(CDL_CV79 == 0), np.count_nonzero(CDL_CV79 == 11))\n",
    "sensitivity79_y = np.divide(np.count_nonzero(CDL_CV79 == 0),denom_sens_79)\n",
    "denom_spec_79 = np.add(np.count_nonzero(CDL_CV79 == 10)- cdl111, np.count_nonzero(CDL_CV79 == -1))\n",
    "specificity79_x = np.divide(np.count_nonzero(CDL_CV79 == 10)- cdl111,denom_spec_79)\n",
    "sub_specificity79_x = np.subtract(1, specificity79_x)\n",
    "\n",
    "#0.8 sensitivity and specificity \n",
    "denom_sens_8 = np.add(np.count_nonzero(CDL_CV8 == 0), np.count_nonzero(CDL_CV8 == 11))\n",
    "sensitivity8_y = np.divide(np.count_nonzero(CDL_CV8 == 0),denom_sens_8)\n",
    "denom_spec_8 = np.add(np.count_nonzero(CDL_CV8 == 10)- cdl111, np.count_nonzero(CDL_CV8 == -1))\n",
    "specificity8_x = np.divide(np.count_nonzero(CDL_CV8 == 10)- cdl111,denom_spec_8)\n",
    "sub_specificity8_x = np.subtract(1, specificity8_x)\n",
    "\n",
    "denom_sens_81 = np.add(np.count_nonzero(CDL_CV81 == 0), np.count_nonzero(CDL_CV81 == 11))\n",
    "sensitivity81_y = np.divide(np.count_nonzero(CDL_CV81 == 0),denom_sens_81)\n",
    "denom_spec_81 = np.add(np.count_nonzero(CDL_CV81 == 10)- cdl111, np.count_nonzero(CDL_CV81 == -1))\n",
    "specificity81_x = np.divide(np.count_nonzero(CDL_CV81 == 10)- cdl111,denom_spec_81)\n",
    "sub_specificity81_x = np.subtract(1, specificity81_x)\n",
    "\n",
    "denom_sens_82 = np.add(np.count_nonzero(CDL_CV82 == 0), np.count_nonzero(CDL_CV82 == 11))\n",
    "sensitivity82_y = np.divide(np.count_nonzero(CDL_CV82 == 0),denom_sens_82)\n",
    "denom_spec_82 = np.add(np.count_nonzero(CDL_CV82 == 10)- cdl111, np.count_nonzero(CDL_CV82 == -1))\n",
    "specificity82_x = np.divide(np.count_nonzero(CDL_CV82 == 10)- cdl111,denom_spec_82)\n",
    "sub_specificity82_x = np.subtract(1, specificity82_x)\n",
    "\n",
    "denom_sens_83 = np.add(np.count_nonzero(CDL_CV83 == 0), np.count_nonzero(CDL_CV83 == 11))\n",
    "sensitivity83_y = np.divide(np.count_nonzero(CDL_CV83 == 0),denom_sens_83)\n",
    "denom_spec_83 = np.add(np.count_nonzero(CDL_CV83 == 10)- cdl111, np.count_nonzero(CDL_CV83 == -1))\n",
    "specificity83_x = np.divide(np.count_nonzero(CDL_CV83 == 10)- cdl111,denom_spec_83)\n",
    "sub_specificity83_x = np.subtract(1, specificity83_x)\n",
    "\n",
    "denom_sens_84 = np.add(np.count_nonzero(CDL_CV84 == 0), np.count_nonzero(CDL_CV84 == 11))\n",
    "sensitivity84_y = np.divide(np.count_nonzero(CDL_CV84 == 0),denom_sens_84)\n",
    "denom_spec_84 = np.add(np.count_nonzero(CDL_CV84 == 10)- cdl111, np.count_nonzero(CDL_CV84 == -1))\n",
    "specificity84_x = np.divide(np.count_nonzero(CDL_CV84 == 10)- cdl111,denom_spec_84)\n",
    "sub_specificity84_x = np.subtract(1, specificity84_x)\n",
    "\n",
    "denom_sens_85 = np.add(np.count_nonzero(CDL_CV85 == 0), np.count_nonzero(CDL_CV85 == 11))\n",
    "sensitivity85_y = np.divide(np.count_nonzero(CDL_CV85 == 0),denom_sens_85)\n",
    "denom_spec_85 = np.add(np.count_nonzero(CDL_CV85 == 10)- cdl111, np.count_nonzero(CDL_CV85 == -1))\n",
    "specificity85_x = np.divide(np.count_nonzero(CDL_CV85 == 10)- cdl111,denom_spec_85)\n",
    "sub_specificity85_x = np.subtract(1, specificity85_x)\n",
    "\n",
    "denom_sens_86 = np.add(np.count_nonzero(CDL_CV86 == 0), np.count_nonzero(CDL_CV86 == 11))\n",
    "sensitivity86_y = np.divide(np.count_nonzero(CDL_CV86 == 0),denom_sens_86)\n",
    "denom_spec_86 = np.add(np.count_nonzero(CDL_CV86 == 10)- cdl111, np.count_nonzero(CDL_CV86 == -1))\n",
    "specificity86_x = np.divide(np.count_nonzero(CDL_CV86 == 10)- cdl111,denom_spec_86)\n",
    "sub_specificity86_x = np.subtract(1, specificity86_x)\n",
    "\n",
    "denom_sens_87 = np.add(np.count_nonzero(CDL_CV87 == 0), np.count_nonzero(CDL_CV87 == 11))\n",
    "sensitivity87_y = np.divide(np.count_nonzero(CDL_CV87 == 0),denom_sens_87)\n",
    "denom_spec_87 = np.add(np.count_nonzero(CDL_CV87 == 10)- cdl111, np.count_nonzero(CDL_CV87 == -1))\n",
    "specificity87_x = np.divide(np.count_nonzero(CDL_CV87 == 10)- cdl111,denom_spec_87)\n",
    "sub_specificity87_x = np.subtract(1, specificity87_x)\n",
    "\n",
    "denom_sens_88 = np.add(np.count_nonzero(CDL_CV88 == 0), np.count_nonzero(CDL_CV88 == 11))\n",
    "sensitivity88_y = np.divide(np.count_nonzero(CDL_CV88 == 0),denom_sens_88)\n",
    "denom_spec_88 = np.add(np.count_nonzero(CDL_CV88 == 10)- cdl111, np.count_nonzero(CDL_CV88 == -1))\n",
    "specificity88_x = np.divide(np.count_nonzero(CDL_CV88 == 10)- cdl111,denom_spec_88)\n",
    "sub_specificity88_x = np.subtract(1, specificity88_x)\n",
    "\n",
    "denom_sens_89 = np.add(np.count_nonzero(CDL_CV89 == 0), np.count_nonzero(CDL_CV89 == 11))\n",
    "sensitivity89_y = np.divide(np.count_nonzero(CDL_CV89 == 0),denom_sens_89)\n",
    "denom_spec_89 = np.add(np.count_nonzero(CDL_CV89 == 10)- cdl111, np.count_nonzero(CDL_CV89 == -1))\n",
    "specificity89_x = np.divide(np.count_nonzero(CDL_CV89 == 10)- cdl111,denom_spec_89)\n",
    "sub_specificity89_x = np.subtract(1, specificity89_x)\n",
    "\n",
    "#0.9 sensitivity and specificity \n",
    "denom_sens_9 = np.add(np.count_nonzero(CDL_CV9 == 0), np.count_nonzero(CDL_CV9 == 11))\n",
    "sensitivity9_y = np.divide(np.count_nonzero(CDL_CV9 == 0),denom_sens_9)\n",
    "denom_spec_9 = np.add(np.count_nonzero(CDL_CV9 == 10)- cdl111, np.count_nonzero(CDL_CV9 == -1))\n",
    "specificity9_x = np.divide(np.count_nonzero(CDL_CV9 == 10)- cdl111,denom_spec_9)\n",
    "sub_specificity9_x = np.subtract(1, specificity9_x)\n",
    "\n",
    "denom_sens_91 = np.add(np.count_nonzero(CDL_CV91 == 0), np.count_nonzero(CDL_CV91 == 11))\n",
    "sensitivity91_y = np.divide(np.count_nonzero(CDL_CV91 == 0),denom_sens_91)\n",
    "denom_spec_91 = np.add(np.count_nonzero(CDL_CV91 == 10)- cdl111, np.count_nonzero(CDL_CV91 == -1))\n",
    "specificity91_x = np.divide(np.count_nonzero(CDL_CV91 == 10)- cdl111,denom_spec_91)\n",
    "sub_specificity91_x = np.subtract(1, specificity91_x)\n",
    "\n",
    "denom_sens_92 = np.add(np.count_nonzero(CDL_CV92 == 0), np.count_nonzero(CDL_CV92 == 11))\n",
    "sensitivity92_y = np.divide(np.count_nonzero(CDL_CV92 == 0),denom_sens_92)\n",
    "denom_spec_92 = np.add(np.count_nonzero(CDL_CV92 == 10)- cdl111, np.count_nonzero(CDL_CV92 == -1))\n",
    "specificity92_x = np.divide(np.count_nonzero(CDL_CV92 == 10)- cdl111,denom_spec_92)\n",
    "sub_specificity92_x = np.subtract(1, specificity92_x)\n",
    "\n",
    "denom_sens_93 = np.add(np.count_nonzero(CDL_CV93 == 0), np.count_nonzero(CDL_CV93 == 11))\n",
    "sensitivity93_y = np.divide(np.count_nonzero(CDL_CV93 == 0),denom_sens_93)\n",
    "denom_spec_93 = np.add(np.count_nonzero(CDL_CV93 == 10)- cdl111, np.count_nonzero(CDL_CV93 == -1))\n",
    "specificity93_x = np.divide(np.count_nonzero(CDL_CV93 == 10)- cdl111,denom_spec_93)\n",
    "sub_specificity93_x = np.subtract(1, specificity93_x)\n",
    "\n",
    "denom_sens_94 = np.add(np.count_nonzero(CDL_CV94 == 0), np.count_nonzero(CDL_CV94 == 11))\n",
    "sensitivity94_y = np.divide(np.count_nonzero(CDL_CV94 == 0),denom_sens_94)\n",
    "denom_spec_94 = np.add(np.count_nonzero(CDL_CV94 == 10)- cdl111, np.count_nonzero(CDL_CV94 == -1))\n",
    "specificity94_x = np.divide(np.count_nonzero(CDL_CV94 == 10)- cdl111,denom_spec_94)\n",
    "sub_specificity94_x = np.subtract(1, specificity94_x)\n",
    "\n",
    "denom_sens_95 = np.add(np.count_nonzero(CDL_CV95 == 0), np.count_nonzero(CDL_CV95 == 11))\n",
    "sensitivity95_y = np.divide(np.count_nonzero(CDL_CV95 == 0),denom_sens_95)\n",
    "denom_spec_95 = np.add(np.count_nonzero(CDL_CV95 == 10)- cdl111, np.count_nonzero(CDL_CV95 == -1))\n",
    "specificity95_x = np.divide(np.count_nonzero(CDL_CV95 == 10)- cdl111,denom_spec_95)\n",
    "sub_specificity95_x = np.subtract(1, specificity95_x)\n",
    "\n",
    "denom_sens_96 = np.add(np.count_nonzero(CDL_CV96 == 0), np.count_nonzero(CDL_CV96 == 11))\n",
    "sensitivity96_y = np.divide(np.count_nonzero(CDL_CV96 == 0),denom_sens_96)\n",
    "denom_spec_96 = np.add(np.count_nonzero(CDL_CV96 == 10)- cdl111, np.count_nonzero(CDL_CV96 == -1))\n",
    "specificity96_x = np.divide(np.count_nonzero(CDL_CV96 == 10)- cdl111,denom_spec_96)\n",
    "sub_specificity96_x = np.subtract(1, specificity96_x)\n",
    "\n",
    "denom_sens_97 = np.add(np.count_nonzero(CDL_CV97 == 0), np.count_nonzero(CDL_CV97 == 11))\n",
    "sensitivity97_y = np.divide(np.count_nonzero(CDL_CV97 == 0),denom_sens_97)\n",
    "denom_spec_97 = np.add(np.count_nonzero(CDL_CV97 == 10)- cdl111, np.count_nonzero(CDL_CV97 == -1))\n",
    "specificity97_x = np.divide(np.count_nonzero(CDL_CV97 == 10)- cdl111,denom_spec_97)\n",
    "sub_specificity97_x = np.subtract(1, specificity97_x)\n",
    "\n",
    "denom_sens_98 = np.add(np.count_nonzero(CDL_CV98 == 0), np.count_nonzero(CDL_CV98 == 11))\n",
    "sensitivity98_y = np.divide(np.count_nonzero(CDL_CV98 == 0),denom_sens_98)\n",
    "denom_spec_98 = np.add(np.count_nonzero(CDL_CV98 == 10)- cdl111, np.count_nonzero(CDL_CV98 == -1))\n",
    "specificity98_x = np.divide(np.count_nonzero(CDL_CV98 == 10)- cdl111,denom_spec_98)\n",
    "sub_specificity98_x = np.subtract(1, specificity98_x)\n",
    "\n",
    "denom_sens_99 = np.add(np.count_nonzero(CDL_CV99 == 0), np.count_nonzero(CDL_CV99 == 11))\n",
    "sensitivity99_y = np.divide(np.count_nonzero(CDL_CV99 == 0),denom_sens_99)\n",
    "denom_spec_99 = np.add(np.count_nonzero(CDL_CV99 == 10)- cdl111, np.count_nonzero(CDL_CV99 == -1))\n",
    "specificity99_x = np.divide(np.count_nonzero(CDL_CV99 == 10)- cdl111,denom_spec_99)\n",
    "sub_specificity99_x = np.subtract(1, specificity99_x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Visualizing the ROC Curve"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Ploting 1 - Specificity is plotted by Sensitivity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "data55 = np.array ([\n",
    "        [sub_specificity0_x, sensitivity0_y],\n",
    "        [sub_specificity01_x, sensitivity01_y],\n",
    "        [sub_specificity02_x, sensitivity02_y],\n",
    "        [sub_specificity03_x, sensitivity03_y],\n",
    "        [sub_specificity04_x, sensitivity04_y],\n",
    "        [sub_specificity05_x, sensitivity05_y],\n",
    "        [sub_specificity06_x, sensitivity06_y],\n",
    "        [sub_specificity07_x, sensitivity07_y],\n",
    "        [sub_specificity08_x, sensitivity08_y],\n",
    "        [sub_specificity09_x, sensitivity09_y],\n",
    "        [sub_specificity1_x, sensitivity1_y],\n",
    "        [sub_specificity11_x, sensitivity11_y],\n",
    "        [sub_specificity12_x, sensitivity12_y],\n",
    "        [sub_specificity13_x, sensitivity13_y],\n",
    "        [sub_specificity14_x, sensitivity14_y],\n",
    "        [sub_specificity15_x, sensitivity15_y],\n",
    "        [sub_specificity16_x, sensitivity16_y],\n",
    "        [sub_specificity17_x, sensitivity17_y],\n",
    "        [sub_specificity18_x, sensitivity18_y],\n",
    "        [sub_specificity19_x, sensitivity19_y],\n",
    "        [sub_specificity2_x, sensitivity2_y],\n",
    "        [sub_specificity21_x, sensitivity21_y],\n",
    "        [sub_specificity22_x, sensitivity22_y],\n",
    "        [sub_specificity23_x, sensitivity23_y],\n",
    "        [sub_specificity24_x, sensitivity24_y],\n",
    "        [sub_specificity25_x, sensitivity25_y],\n",
    "        [sub_specificity26_x, sensitivity26_y],\n",
    "        [sub_specificity27_x, sensitivity27_y],\n",
    "        [sub_specificity28_x, sensitivity28_y],\n",
    "        [sub_specificity29_x, sensitivity29_y],\n",
    "        [sub_specificity3_x, sensitivity3_y],\n",
    "        [sub_specificity31_x, sensitivity31_y],\n",
    "        [sub_specificity32_x, sensitivity32_y],\n",
    "        [sub_specificity33_x, sensitivity33_y],\n",
    "        [sub_specificity34_x, sensitivity34_y],\n",
    "        [sub_specificity35_x, sensitivity35_y],\n",
    "        [sub_specificity36_x, sensitivity36_y],\n",
    "        [sub_specificity37_x, sensitivity37_y],\n",
    "        [sub_specificity38_x, sensitivity38_y],\n",
    "        [sub_specificity39_x, sensitivity39_y],\n",
    "        [sub_specificity4_x, sensitivity4_y],\n",
    "        [sub_specificity41_x, sensitivity41_y],\n",
    "        [sub_specificity42_x, sensitivity42_y],\n",
    "        [sub_specificity43_x, sensitivity43_y],\n",
    "        [sub_specificity44_x, sensitivity44_y],\n",
    "        [sub_specificity45_x, sensitivity45_y],\n",
    "        [sub_specificity46_x, sensitivity46_y],\n",
    "        [sub_specificity47_x, sensitivity47_y],\n",
    "        [sub_specificity48_x, sensitivity48_y],\n",
    "        [sub_specificity49_x, sensitivity49_y],\n",
    "        [sub_specificity5_x, sensitivity5_y],\n",
    "        [sub_specificity51_x, sensitivity51_y],\n",
    "        [sub_specificity52_x, sensitivity52_y],\n",
    "        [sub_specificity53_x, sensitivity53_y],\n",
    "        [sub_specificity54_x, sensitivity54_y],\n",
    "        [sub_specificity55_x, sensitivity55_y],\n",
    "        [sub_specificity56_x, sensitivity56_y],\n",
    "        [sub_specificity57_x, sensitivity57_y],\n",
    "        [sub_specificity58_x, sensitivity58_y],\n",
    "        [sub_specificity59_x, sensitivity59_y],\n",
    "        [sub_specificity6_x, sensitivity6_y],\n",
    "        [sub_specificity61_x, sensitivity61_y],\n",
    "        [sub_specificity62_x, sensitivity62_y],\n",
    "        [sub_specificity63_x, sensitivity63_y],\n",
    "        [sub_specificity64_x, sensitivity64_y],\n",
    "        [sub_specificity65_x, sensitivity65_y],\n",
    "        [sub_specificity66_x, sensitivity66_y],\n",
    "        [sub_specificity67_x, sensitivity67_y],\n",
    "        [sub_specificity68_x, sensitivity68_y],\n",
    "        [sub_specificity69_x, sensitivity69_y],\n",
    "        [sub_specificity7_x, sensitivity7_y],\n",
    "        [sub_specificity71_x, sensitivity71_y],\n",
    "        [sub_specificity72_x, sensitivity72_y],\n",
    "        [sub_specificity73_x, sensitivity73_y],\n",
    "        [sub_specificity74_x, sensitivity74_y],\n",
    "        [sub_specificity75_x, sensitivity75_y],\n",
    "        [sub_specificity76_x, sensitivity76_y],\n",
    "        [sub_specificity77_x, sensitivity77_y],\n",
    "        [sub_specificity78_x, sensitivity78_y],\n",
    "        [sub_specificity79_x, sensitivity79_y],\n",
    "        [sub_specificity8_x, sensitivity8_y],\n",
    "        [sub_specificity81_x, sensitivity81_y],\n",
    "        [sub_specificity82_x, sensitivity82_y],\n",
    "        [sub_specificity83_x, sensitivity83_y],\n",
    "        [sub_specificity84_x, sensitivity84_y],\n",
    "        [sub_specificity85_x, sensitivity85_y],\n",
    "        [sub_specificity86_x, sensitivity86_y],\n",
    "        [sub_specificity87_x, sensitivity87_y],\n",
    "        [sub_specificity88_x, sensitivity88_y],\n",
    "        [sub_specificity89_x, sensitivity89_y], \n",
    "        [sub_specificity9_x, sensitivity9_y],\n",
    "        [sub_specificity91_x, sensitivity91_y],\n",
    "        [sub_specificity92_x, sensitivity92_y],\n",
    "        [sub_specificity93_x, sensitivity93_y],\n",
    "        [sub_specificity94_x, sensitivity94_y],\n",
    "        [sub_specificity95_x, sensitivity95_y],\n",
    "        [sub_specificity96_x, sensitivity96_y],\n",
    "        [sub_specificity97_x, sensitivity97_y],\n",
    "        [sub_specificity98_x, sensitivity98_y],\n",
    "        [sub_specificity99_x, sensitivity99_y], \n",
    "])\n",
    "data55\n",
    "x, y = data55.T\n",
    "plt.plot(x, y, '-')\n",
    "plt.plot([0.0, 1.0], [0.0,1.0], linestyle= \"--\", color = \"red\")\n",
    "\n",
    "#arrows highlighting where each threshold value is\n",
    "plt.annotate('0.1 Threshold', xy=(sub_specificity1_x, sensitivity1_y),arrowprops=dict(arrowstyle='->'), xytext=(0.87, 0.85)) \n",
    "plt.annotate('0.2 Threshold', xy=(sub_specificity2_x, sensitivity2_y),arrowprops=dict(arrowstyle='->'), xytext=(0.75, 0.9)) \n",
    "plt.annotate('0.3 Threshold', xy=(sub_specificity3_x, sensitivity3_y),arrowprops=dict(arrowstyle='->'), xytext=(0.4, 0.83)) \n",
    "plt.annotate('0.4 Threshold', xy=(sub_specificity4_x, sensitivity4_y),arrowprops=dict(arrowstyle='->'), xytext=(0.05, 0.7)) \n",
    "plt.annotate('0.5 Threshold', xy=(sub_specificity5_x, sensitivity5_y),arrowprops=dict(arrowstyle='->'), xytext=(0.2, 0.65)) \n",
    "plt.annotate('0.6 Threshold', xy=(sub_specificity6_x, sensitivity6_y),arrowprops=dict(arrowstyle='->'), xytext=(0.2, 0.5)) \n",
    "plt.annotate('0.7 Threshold', xy=(sub_specificity7_x, sensitivity7_y),arrowprops=dict(arrowstyle='->'), xytext=(0.25, 0.4)) \n",
    "plt.annotate('0.8 Threshold', xy=(sub_specificity8_x, sensitivity8_y),arrowprops=dict(arrowstyle='->'), xytext=(0.3, 0.2)) \n",
    "plt.annotate('0.9 Threshold', xy=(sub_specificity9_x, sensitivity9_y),arrowprops=dict(arrowstyle='->'), xytext=(0.1, 0.05)) \n",
    "\n",
    "#Setting up plot layout\n",
    "plt.ylim(0, 1)\n",
    "plt.xlim(-0.01, 1)\n",
    "plt.grid()\n",
    "plt.title('Reciever Operating Characteristic')\n",
    "plt.ylabel('True Positive Rate (sensitivity)', fontsize=12)\n",
    "plt.xlabel('False Positive Rate (1-specificity)', fontsize=12)\n",
    "plt.rcParams['figure.figsize'] = (10,10)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Using Youden's Index to Find the Ideal Classification Threshold"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Youden's Index is the point on the ROC curve where sensitivity and specificity are maximized, a maximum difference between TPR and FPR (TPR - FPR). The threshold ranges between 0 and 1. A plot is generated to show the point on the curve where the Youden's Index point is on curve."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#setting the inputs\n",
    "\n",
    "#100 different CV thresholds used \n",
    "thresholds = [0.00, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.2, 0.21, 0.22, 0.23, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.30, 0.31, 0.32, 0.33, 0.34, 0.35, 0.36, 0.37, 0.38, 0.39, 0.4, 0.41, 0.42, 0.43, 0.44, 0.45, 0.46, 0.47, 0.48, 0.49, 0.50, 0.51, 0.52, 0.53, 0.54, 0.55, 0.56, 0.57, 0.58, 0.59, 0.6, 0.61, 0.62, 0.63, 0.64, 0.65, 0.66, 0.67, 0.68, 0.69, 0.7, 0.71, 0.72, 0.73, 0.74, 0.75, 0.76, 0.77, 0.78, 0.79, 0.80, 0.81, 0.82, 0.83, 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.9, 0.91, 0.92, 0.93, 0.94, 0.95, 0.96, 0.97, 0.98, 0.99]\n",
    "fpr = np.array([sub_specificity0_x, sub_specificity01_x, sub_specificity02_x, sub_specificity03_x, sub_specificity04_x, sub_specificity05_x, sub_specificity06_x, sub_specificity07_x, sub_specificity08_x, sub_specificity09_x, sub_specificity1_x, sub_specificity11_x, sub_specificity12_x, sub_specificity13_x, sub_specificity14_x, sub_specificity15_x, sub_specificity16_x, sub_specificity17_x, sub_specificity18_x, sub_specificity19_x, sub_specificity2_x, sub_specificity21_x, sub_specificity22_x, sub_specificity23_x, sub_specificity24_x, sub_specificity25_x, sub_specificity26_x, sub_specificity27_x, sub_specificity28_x, sub_specificity29_x, sub_specificity3_x, sub_specificity31_x, sub_specificity32_x, sub_specificity33_x, sub_specificity34_x, sub_specificity35_x, sub_specificity36_x, sub_specificity37_x, sub_specificity38_x, sub_specificity39_x, sub_specificity4_x, sub_specificity41_x, sub_specificity42_x, sub_specificity43_x, sub_specificity44_x, sub_specificity45_x, sub_specificity46_x, sub_specificity47_x, sub_specificity48_x, sub_specificity49_x, sub_specificity5_x, sub_specificity51_x, sub_specificity52_x, sub_specificity53_x, sub_specificity54_x, sub_specificity55_x, sub_specificity56_x, sub_specificity57_x, sub_specificity58_x, sub_specificity59_x, sub_specificity6_x, sub_specificity61_x, sub_specificity62_x, sub_specificity63_x, sub_specificity64_x, sub_specificity65_x, sub_specificity66_x, sub_specificity67_x, sub_specificity68_x, sub_specificity69_x, sub_specificity7_x, sub_specificity71_x, sub_specificity72_x, sub_specificity73_x, sub_specificity74_x, sub_specificity75_x, sub_specificity76_x, sub_specificity77_x, sub_specificity78_x, sub_specificity79_x, sub_specificity8_x, sub_specificity81_x, sub_specificity82_x, sub_specificity83_x, sub_specificity84_x, sub_specificity85_x, sub_specificity86_x, sub_specificity87_x, sub_specificity88_x, sub_specificity89_x, sub_specificity9_x, sub_specificity91_x, sub_specificity92_x, sub_specificity93_x, sub_specificity94_x, sub_specificity95_x, sub_specificity96_x, sub_specificity97_x, sub_specificity98_x, sub_specificity99_x])\n",
    "tpr = np.array([sensitivity0_y, sensitivity01_y, sensitivity02_y, sensitivity03_y, sensitivity04_y, sensitivity05_y, sensitivity06_y, sensitivity07_y, sensitivity08_y, sensitivity09_y, sensitivity1_y, sensitivity11_y, sensitivity12_y, sensitivity13_y, sensitivity14_y, sensitivity15_y, sensitivity16_y, sensitivity17_y, sensitivity18_y, sensitivity19_y, sensitivity2_y, sensitivity21_y, sensitivity22_y, sensitivity23_y, sensitivity24_y, sensitivity25_y, sensitivity26_y, sensitivity27_y, sensitivity28_y, sensitivity29_y, sensitivity3_y, sensitivity31_y, sensitivity32_y, sensitivity33_y, sensitivity34_y, sensitivity35_y, sensitivity36_y, sensitivity37_y, sensitivity38_y, sensitivity39_y, sensitivity4_y, sensitivity41_y, sensitivity42_y, sensitivity43_y, sensitivity44_y, sensitivity45_y, sensitivity46_y, sensitivity47_y, sensitivity48_y, sensitivity49_y, sensitivity5_y, sensitivity51_y, sensitivity52_y, sensitivity53_y, sensitivity54_y, sensitivity55_y, sensitivity56_y, sensitivity57_y, sensitivity58_y, sensitivity59_y, sensitivity6_y, sensitivity61_y, sensitivity62_y, sensitivity63_y, sensitivity64_y, sensitivity65_y, sensitivity66_y, sensitivity67_y, sensitivity68_y, sensitivity69_y, sensitivity7_y, sensitivity71_y, sensitivity72_y, sensitivity73_y, sensitivity74_y, sensitivity75_y, sensitivity76_y, sensitivity77_y, sensitivity78_y, sensitivity79_y, sensitivity8_y, sensitivity81_y, sensitivity82_y, sensitivity83_y, sensitivity84_y, sensitivity85_y, sensitivity86_y, sensitivity87_y, sensitivity88_y, sensitivity89_y, sensitivity9_y, sensitivity91_y, sensitivity92_y, sensitivity93_y, sensitivity94_y, sensitivity95_y, sensitivity96_y, sensitivity97_y, sensitivity98_y, sensitivity99_y])\n",
    "\n",
    "#finding the threshold value with the greatest difference between TPR and FPR (x and y axes)\n",
    "def cutoff_youdens_j(fpr,tpr,thresholds):\n",
    "    j_scores = tpr-fpr\n",
    "    j_ordered = sorted(zip(j_scores,thresholds))\n",
    "    return j_ordered[-1][1]\n",
    "\n",
    "best_thresh = cutoff_youdens_j(fpr,tpr,thresholds)\n",
    "print (best_thresh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#displays Youden's Index calculated point on cure \n",
    "\n",
    "best_thresh_index = thresholds.index(best_thresh)\n",
    "fpr_specificity_ideal = fpr[best_thresh_index] \n",
    "tpr_sensitivity_ideal = tpr[best_thresh_index]\n",
    "\n",
    "data100 = np.array ([\n",
    "        [sub_specificity0_x, sensitivity0_y],\n",
    "        [sub_specificity01_x, sensitivity01_y],\n",
    "        [sub_specificity02_x, sensitivity02_y],\n",
    "        [sub_specificity03_x, sensitivity03_y],\n",
    "        [sub_specificity04_x, sensitivity04_y],\n",
    "        [sub_specificity05_x, sensitivity05_y],\n",
    "        [sub_specificity06_x, sensitivity06_y],\n",
    "        [sub_specificity07_x, sensitivity07_y],\n",
    "        [sub_specificity08_x, sensitivity08_y],\n",
    "        [sub_specificity09_x, sensitivity09_y],\n",
    "        [sub_specificity1_x, sensitivity1_y],\n",
    "        [sub_specificity11_x, sensitivity11_y],\n",
    "        [sub_specificity12_x, sensitivity12_y],\n",
    "        [sub_specificity13_x, sensitivity13_y],\n",
    "        [sub_specificity14_x, sensitivity14_y],\n",
    "        [sub_specificity15_x, sensitivity15_y],\n",
    "        [sub_specificity16_x, sensitivity16_y],\n",
    "        [sub_specificity17_x, sensitivity17_y],\n",
    "        [sub_specificity18_x, sensitivity18_y],\n",
    "        [sub_specificity19_x, sensitivity19_y],\n",
    "        [sub_specificity2_x, sensitivity2_y],\n",
    "        [sub_specificity21_x, sensitivity21_y],\n",
    "        [sub_specificity22_x, sensitivity22_y],\n",
    "        [sub_specificity23_x, sensitivity23_y],\n",
    "        [sub_specificity24_x, sensitivity24_y],\n",
    "        [sub_specificity25_x, sensitivity25_y],\n",
    "        [sub_specificity26_x, sensitivity26_y],\n",
    "        [sub_specificity27_x, sensitivity27_y],\n",
    "        [sub_specificity28_x, sensitivity28_y],\n",
    "        [sub_specificity29_x, sensitivity29_y],\n",
    "        [sub_specificity3_x, sensitivity3_y],\n",
    "        [sub_specificity31_x, sensitivity31_y],\n",
    "        [sub_specificity32_x, sensitivity32_y],\n",
    "        [sub_specificity33_x, sensitivity33_y],\n",
    "        [sub_specificity34_x, sensitivity34_y],\n",
    "        [sub_specificity35_x, sensitivity35_y],\n",
    "        [sub_specificity36_x, sensitivity36_y],\n",
    "        [sub_specificity37_x, sensitivity37_y],\n",
    "        [sub_specificity38_x, sensitivity38_y],\n",
    "        [sub_specificity39_x, sensitivity39_y],\n",
    "        [sub_specificity4_x, sensitivity4_y],\n",
    "        [sub_specificity41_x, sensitivity41_y],\n",
    "        [sub_specificity42_x, sensitivity42_y],\n",
    "        [sub_specificity43_x, sensitivity43_y],\n",
    "        [sub_specificity44_x, sensitivity44_y],\n",
    "        [sub_specificity45_x, sensitivity45_y],\n",
    "        [sub_specificity46_x, sensitivity46_y],\n",
    "        [sub_specificity47_x, sensitivity47_y],\n",
    "        [sub_specificity48_x, sensitivity48_y],\n",
    "        [sub_specificity49_x, sensitivity49_y],\n",
    "        [sub_specificity5_x, sensitivity5_y],\n",
    "        [sub_specificity51_x, sensitivity51_y],\n",
    "        [sub_specificity52_x, sensitivity52_y],\n",
    "        [sub_specificity53_x, sensitivity53_y],\n",
    "        [sub_specificity54_x, sensitivity54_y],\n",
    "        [sub_specificity55_x, sensitivity55_y],\n",
    "        [sub_specificity56_x, sensitivity56_y],\n",
    "        [sub_specificity57_x, sensitivity57_y],\n",
    "        [sub_specificity58_x, sensitivity58_y],\n",
    "        [sub_specificity59_x, sensitivity59_y],\n",
    "        [sub_specificity6_x, sensitivity6_y],\n",
    "        [sub_specificity61_x, sensitivity61_y],\n",
    "        [sub_specificity62_x, sensitivity62_y],\n",
    "        [sub_specificity63_x, sensitivity63_y],\n",
    "        [sub_specificity64_x, sensitivity64_y],\n",
    "        [sub_specificity65_x, sensitivity65_y],\n",
    "        [sub_specificity66_x, sensitivity66_y],\n",
    "        [sub_specificity67_x, sensitivity67_y],\n",
    "        [sub_specificity68_x, sensitivity68_y],\n",
    "        [sub_specificity69_x, sensitivity69_y],\n",
    "        [sub_specificity7_x, sensitivity7_y],\n",
    "        [sub_specificity71_x, sensitivity71_y],\n",
    "        [sub_specificity72_x, sensitivity72_y],\n",
    "        [sub_specificity73_x, sensitivity73_y],\n",
    "        [sub_specificity74_x, sensitivity74_y],\n",
    "        [sub_specificity75_x, sensitivity75_y],\n",
    "        [sub_specificity76_x, sensitivity76_y],\n",
    "        [sub_specificity77_x, sensitivity77_y],\n",
    "        [sub_specificity78_x, sensitivity78_y],\n",
    "        [sub_specificity79_x, sensitivity79_y],\n",
    "        [sub_specificity8_x, sensitivity8_y],\n",
    "        [sub_specificity81_x, sensitivity81_y],\n",
    "        [sub_specificity82_x, sensitivity82_y],\n",
    "        [sub_specificity83_x, sensitivity83_y],\n",
    "        [sub_specificity84_x, sensitivity84_y],\n",
    "        [sub_specificity85_x, sensitivity85_y],\n",
    "        [sub_specificity86_x, sensitivity86_y],\n",
    "        [sub_specificity87_x, sensitivity87_y],\n",
    "        [sub_specificity88_x, sensitivity88_y],\n",
    "        [sub_specificity89_x, sensitivity89_y], \n",
    "        [sub_specificity9_x, sensitivity9_y],\n",
    "        [sub_specificity91_x, sensitivity91_y],\n",
    "        [sub_specificity92_x, sensitivity92_y],\n",
    "        [sub_specificity93_x, sensitivity93_y],\n",
    "        [sub_specificity94_x, sensitivity94_y],\n",
    "        [sub_specificity95_x, sensitivity95_y],\n",
    "        [sub_specificity96_x, sensitivity96_y],\n",
    "        [sub_specificity97_x, sensitivity97_y],\n",
    "        [sub_specificity98_x, sensitivity98_y],\n",
    "        [sub_specificity99_x, sensitivity99_y], \n",
    "])\n",
    "data100\n",
    "x, y = data100.T\n",
    "plt.scatter(x, y)\n",
    "plt.plot([0.0, 1.0], [0.0,1.0], linestyle= \"--\", color = \"red\")\n",
    "plt.plot([fpr_specificity_ideal], [tpr_sensitivity_ideal], marker='o', markersize=8, color=\"red\")\n",
    "\n",
    "#arrows highlighting where each threshold value is\n",
    "plt.annotate('0.1 Threshold', xy=(sub_specificity1_x, sensitivity1_y),arrowprops=dict(arrowstyle='->'), xytext=(0.87, 0.85)) \n",
    "plt.annotate('0.2 Threshold', xy=(sub_specificity2_x, sensitivity2_y),arrowprops=dict(arrowstyle='->'), xytext=(0.81, 0.94)) \n",
    "plt.annotate('0.3 Threshold', xy=(sub_specificity3_x, sensitivity3_y),arrowprops=dict(arrowstyle='->'), xytext=(0.67, 0.93)) \n",
    "plt.annotate('0.4 Threshold', xy=(sub_specificity4_x, sensitivity4_y),arrowprops=dict(arrowstyle='->'), xytext=(0.61, 0.88)) \n",
    "plt.annotate('0.5 Threshold', xy=(sub_specificity5_x, sensitivity5_y),arrowprops=dict(arrowstyle='->'), xytext=(0.22, 0.75)) \n",
    "plt.annotate('0.6 Threshold', xy=(sub_specificity6_x, sensitivity6_y),arrowprops=dict(arrowstyle='->'), xytext=(0.07, 0.5)) \n",
    "plt.annotate('0.7 Threshold', xy=(sub_specificity7_x, sensitivity7_y),arrowprops=dict(arrowstyle='->'), xytext=(0.07, 0.3)) \n",
    "plt.annotate('0.8 Threshold', xy=(sub_specificity8_x, sensitivity8_y),arrowprops=dict(arrowstyle='->'), xytext=(0.025, 0.17)) \n",
    "plt.annotate('0.9 Threshold', xy=(sub_specificity9_x, sensitivity9_y),arrowprops=dict(arrowstyle='->'), xytext=(0.07, 0.03)) \n",
    "\n",
    "#Setting plot layout\n",
    "plt.grid()\n",
    "plt.title('Reciever Operating Characteristic')\n",
    "plt.ylabel('True Positive Rate (sensitivity)', fontsize=12)\n",
    "plt.xlabel('False Positive Rate (1-specificity)', fontsize=12)\n",
    "plt.rcParams['figure.figsize'] = (10,10)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Finding the AUC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Area Under the ROC curve (AUC) is a performance measurement (between 0 and 1) describing the discrimination power between the crop and non-crop classes. The higher the AUC value the better the model is at separating the two classes. An AUC value of 0.5 signifies there is no distinguishability between the classes. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "AUC = (round(sklearn.metrics.auc(fpr, tpr, reorder= 'deprecated'), 2))\n",
    "print (\"Area Under Curve (AUC):\", AUC)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Generating Classified Image Using Youden's Index Optimal Threshold "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Creates a crop/non-crop clssification based on the Youden's Index threshold and displays the image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#Classify image using the Youden's Index Optimal Threshold\n",
    "CV_reclass_ideal = np.copy(CV)\n",
    "CV_reclass_ideal[np.where(CV_reclass_ideal>= best_thresh)] = 1\n",
    "CV_reclass_ideal[np.where(CV_reclass_ideal< best_thresh)] = 0\n",
    "CVideal_water = np.subtract(CV_reclass_ideal, watermask)\n",
    "CVideal_crop_non = np.copy(CVideal_water)\n",
    "CVideal_crop_non[np.where(CVideal_crop_non>0)] = 1\n",
    "CVideal_crop_non[np.where(CVideal_crop_non<=0)] = -10\n",
    "CDL_CVideal = np.subtract(CDL_binary, CVideal_crop_non)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Displaying binary crop/non-crop classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "arrayLucc = CV_reclass_ideal\n",
    "#define the color map and value name as a dictionary\n",
    "t = 1 ## alpha value\n",
    "cmap = {1:[0.0,1.0,0.0,t],0:[1.0,0.1,0.1,t]}\n",
    "labels = {1:'Crop',0:'Non-crop'}\n",
    "arrayShow = np.array([[cmap[i] for i in j] for j in arrayLucc])    \n",
    "## create patches as legend\n",
    "patches =[mpatches.Patch(color=cmap[i],label=labels[i]) for i in cmap]\n",
    "\n",
    "plt.title('Classified Image', fontsize=15)\n",
    "plt.imshow(arrayShow)\n",
    "plt.legend(handles=patches, bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0. )\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Displaying crop/non-crop classification with mask and CDL accuracy check included."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#For classification display where maked values are \n",
    "\n",
    "CDL_CVideal_with_mask = CDL_CVideal + watermask\n",
    "CDL_CVideal_with_mask_rc = np.copy(CDL_CVideal_with_mask)\n",
    "CDL_CVideal_with_mask_rc[np.where(watermask== 111)]= 100 \n",
    "\n",
    "#arrayLucc is the array of classified classes\n",
    "arrayLucc = CDL_CVideal_with_mask_rc\n",
    "#define the color map and value name as a dictionary\n",
    "t = 1 ## alpha value\n",
    "cmap = {0:[0.0,1.0,0.0,t],11:[1.0,0.5,0.1,t],10:[1.0,0.1,0.1,t],-1:[0,0.3,1.0,t],100:[0,0,0,t]}\n",
    "labels = {0:'Correct Crop',11:'Incorrect Crop',10:'Correct Non-crop',-1:'Incorrect Non-crop',100:'Mask'}\n",
    "arrayShow = np.array([[cmap[i] for i in j] for j in arrayLucc])    \n",
    "## create patches as legend\n",
    "patches =[mpatches.Patch(color=cmap[i],label=labels[i]) for i in cmap]\n",
    "\n",
    "plt.title('Classified Image', fontsize=15)\n",
    "plt.imshow(arrayShow)\n",
    "plt.legend(handles=patches, bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0. )\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Getting Accuracy Statictics for the Youden's Index Optimal Threshold"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "1. Returns the J-statistic for the determined threshold (maximum difference value between TPR and FPR ranging between 0 and 1)\n",
    "2. Get percent crop and non-crop classified correctly and incorrectly\n",
    "3. Get User's and Producer's accuracy of the classification for both crop and non-crop classes\n",
    "4. Calculate the Kapppa Coefficient = the measure of the agreement between the SAR imagery derived classification and the CDL\n",
    "    Kappa = (total pixel x total correct pixel - sum of the products)/(total pixel^2 - sum of products)\n",
    "    Sum of products = sum of row total x column total \n",
    "        Interpreting Kappa Statistic: \n",
    "         < 0.20          = Poor agreement \n",
    "         0.20 - 0.40     = Fair agreement \n",
    "         0.40 - 0.60     = Moderate agreement \n",
    "         0.60 - 0.80     = Good agreement \n",
    "         0.80 - 1.00     = Very good agreement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#Determining the J-statistic calculated by finding the maximum difference between the TPR and FPR\n",
    "j_scores = tpr-fpr\n",
    "j_ordered = sorted(zip(j_scores,thresholds))\n",
    "j_statistic = (round(j_ordered[-1][0], 2))\n",
    "\n",
    "#Getting statistics on the accuracy of the CV classification using Youden's Index optimal threshold based on CDL classifications\n",
    "print ('Statistics of accuracy of using CV to classify crop vs non-crop based on CDL:')\n",
    "p_crop_correct = np.count_nonzero(CDL_CVideal == 0)/(cdl0)\n",
    "p_crop_correct1 = (round((p_crop_correct*100), 2))\n",
    "print ('% correct crop: ',p_crop_correct1)\n",
    "p_non_correct = (np.count_nonzero(CDL_CVideal == 10)-cdl111)/(cdl0)\n",
    "p_non_correct1 = (round((p_non_correct)*100, 2))\n",
    "print ('% correct non-crop: ', p_non_correct1)\n",
    "p_crop_incorrect = np.count_nonzero(CDL_CVideal == 11)/(cdl0)\n",
    "p_crop_incorrect1 = (round((p_crop_incorrect)*100, 2))  \n",
    "print ('% incorrect crop: ', p_crop_incorrect1)\n",
    "p_non_incorrect = np.count_nonzero(CDL_CVideal == -1)/(cdl0)\n",
    "p_non_incorrect1 = (round((p_non_incorrect)*100, 2))    \n",
    "print ('% incorrect non-crop: ', p_non_incorrect1)\n",
    "\n",
    "#Getting Overall Accuracy statistics of the CV classification using Youden's Index optimal threshold based on CDL classifications\n",
    "print ('Statistics of overall accuracy of using CV to classify crop vs non-crop based on CDL')\n",
    "p_overall_correct = round((p_crop_correct + p_non_correct)*100, 2)\n",
    "print ('% overall correct: ', p_overall_correct)\n",
    "p_overall_incorrect = round((p_crop_incorrect + p_non_incorrect)*100, 2)\n",
    "print ('% overall incorrect: ', p_overall_incorrect)\n",
    "\n",
    "#Getting user's and producer's accuracy of the CV classification based on CDL classifications\n",
    "print ('Users and producers accuracy statistics:')\n",
    "crop_correct = np.count_nonzero(CDL_CVideal == 0)\n",
    "non_correct = np.count_nonzero(CDL_CVideal == 10)\n",
    "crop_incorrect = np.count_nonzero(CDL_CVideal == 11)\n",
    "non_incorrect = np.count_nonzero(CDL_CVideal == -1)\n",
    "\n",
    "crop_producers_total = crop_correct + crop_incorrect\n",
    "non_producers_total = (non_correct - cdl111) + non_incorrect\n",
    "\n",
    "crop_users_total = crop_correct + non_incorrect\n",
    "non_users_total = (non_correct - cdl111) + crop_incorrect\n",
    "\n",
    "crop_p_accuracy = round((crop_correct/crop_producers_total)*100, 2)\n",
    "print ('% crop producers accuracy: ', crop_p_accuracy)\n",
    "non_p_accuracy = round(((non_correct - cdl111)/non_producers_total)*100, 2)\n",
    "print('% non-crop producers accuracy: ', non_p_accuracy)\n",
    "\n",
    "crop_u_accuracy = round((crop_correct/crop_users_total)*100, 2)\n",
    "print ('% crop users accuracy: ', crop_u_accuracy)\n",
    "non_u_accuracy = round(((non_correct - cdl111)/non_users_total)*100, 2)\n",
    "print ('% non-crop users accuracy: ', non_u_accuracy)\n",
    "\n",
    "#Calculating Kappa Coefficient\n",
    "Total_correct = np.count_nonzero(CDL_CVideal == 0) + (np.count_nonzero(CDL_CVideal == 10) - cdl111)\n",
    "\n",
    "Total_pixel_count = (cdl0)\n",
    "\n",
    "Sum_of_products_crop_row = np.count_nonzero(CDL_CVideal == 0) + np.count_nonzero(CDL_CVideal == -1)\n",
    "Sum_of_products_non_row = (np.count_nonzero(CDL_CVideal == 10) - cdl111) + np.count_nonzero(CDL_CVideal == 11)\n",
    "Sum_of_products_crop_col = np.count_nonzero(CDL_CVideal == 0) + np.count_nonzero(CDL_CVideal == 11)\n",
    "Sum_of_products_non_col = (np.count_nonzero(CDL_CVideal == 10) - cdl111) + np.count_nonzero(CDL_CVideal == -1)\n",
    "\n",
    "Sum_of_products = (Sum_of_products_crop_row * Sum_of_products_crop_col) + (Sum_of_products_non_row * Sum_of_products_non_col)\n",
    "\n",
    "Kappa_coefficient = round((Total_pixel_count * Total_correct - Sum_of_products) / (Total_pixel_count * Total_pixel_count - Sum_of_products), 2)\n",
    "print ('Kappa Coefficient: ', Kappa_coefficient)\n",
    "\n",
    "if Kappa_coefficient <= 0.2:\n",
    "    print ('Poor agreement')\n",
    "else:\n",
    "    if Kappa_coefficient <= 0.4:\n",
    "        print ('Fair agreement')\n",
    "    else:\n",
    "        if Kappa_coefficient <= 0.6:\n",
    "            print ('Moderate agreement')\n",
    "        else: \n",
    "            if Kappa_coefficient <= 0.8:\n",
    "                print ('Good agreement')\n",
    "            else: \n",
    "                if Kappa_coefficient > 0.8:\n",
    "                    print ('Very good agreement')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Exporting Accuracy Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Writes the calculated accuracy results to an excel CSV file to the set output directory. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "#Exporting accuracy statistics as a CSV file\n",
    "\n",
    "l0 = [best_thresh]\n",
    "l1 = [p_overall_correct]\n",
    "l2 = [p_crop_correct1]\n",
    "l3 = [p_non_correct1]\n",
    "l4 = [p_crop_incorrect1]\n",
    "l5 = [p_non_incorrect1]\n",
    "l6 = [crop_p_accuracy]\n",
    "l7 = [non_p_accuracy]\n",
    "l8 = [crop_u_accuracy]\n",
    "l9 = [non_u_accuracy]\n",
    "l10 = [Kappa_coefficient]\n",
    "l11 = [j_statistic]\n",
    "l12 = [AUC]\n",
    "\n",
    "df = pd.DataFrame({\"Threshold\": l0, \"Overall Correct\": l1, \"% crop correct\": l2, \"% non-crop correct\": l3, \"% crop incorrect\": l4, \"% non-crop incorrect\": l5, \"% Crop Producers Accuracy\": l6, \"% Non-crop Producers Accuarcy\": l7, \"% Crop Users Accuracy\":l8, \"% Non-crop Users Accuracy\": l9, \"Kappa Coefficient\": l10, \"J-statistic\": l11, \"AUC\": l12})\n",
    "\n",
    "os.chdir(Output_folder)\n",
    "\n",
    "df.to_csv(Accuracy_results)\n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "# Exporting the Classified Image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Writes the array to a geotiff that is classified by the Youden's Index ideal threshold based on CV to the set output directory. \n",
    "\n",
    "\n",
    "0 = Crop correct\n",
    "\n",
    "10 = Non-crop correct\n",
    "\n",
    "11 = Crop Incorrect\n",
    "\n",
    "-1 = Non-crop Incorrect\n",
    "\n",
    "(100 = masked values in calcualtion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Define write_geotiff function - writes an array to a geotiff\n",
    "def write_geotiff_export(image, geotrans, cols, rows, spatial_ref, nodata, outfilename):    \n",
    "    driver = gdal.GetDriverByName('GTiff')\n",
    "    outRaster = driver.Create(outfilename, cols, rows, 1, gdal.GDT_Float32)\n",
    "    outRaster.SetGeoTransform(geotrans)\n",
    "    outband = outRaster.GetRasterBand(1)\n",
    "    outband.WriteArray(image)\n",
    "    outRasterSRS = osr.SpatialReference()\n",
    "    outband.SetNoDataValue(np.nan)\n",
    "    outRasterSRS.ImportFromWkt(spatial_ref)\n",
    "    outRaster.SetProjection(outRasterSRS.ExportToWkt())\n",
    "    outband.FlushCache()\n",
    "        \n",
    "os.chdir(Output_folder)\n",
    "\n",
    "#Exports binary classification \n",
    "\n",
    "write_geotiff_export(CV_reclass_ideal, geotransform, cols, rows, spatialref, np.nan, CV_classification_binary)\n",
    "\n",
    "#Exports classification with mask and accuracy with CDL is included\n",
    "\n",
    "write_geotiff_export(CDL_CVideal_with_mask_rc, geotransform, cols, rows, spatialref, np.nan, CV_classification_filename_with_accuracy)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}