Merge pull request #71 from satzbeleg/dev

Bug fixed
ulf1 · Jul 2, 2021 · 5a21137 · 5a21137
2 parents a8b22a5 + 59059d3
commit 5a21137
Show file tree

Hide file tree

Showing 4 changed files with 109 additions and 14 deletions.
diff --git a/README.md b/README.md
@@ -19,7 +19,7 @@ The package `bwsample` addresses three areas:
 The `bwsample` [git repo](http://github.com/satzbeleg/bwsample) is available as [PyPi package](https://pypi.org/project/bwsample)
 
 ```sh
-pip install bwsample>=0.6.5
+pip install bwsample>=0.6.7
 ```
 
 ## Overview

diff --git a/bwsample/__init__.py b/bwsample/__init__.py
@@ -1,4 +1,4 @@
-__version__ = '0.6.6'
+__version__ = '0.6.7'
 
 from .sampling import sample
 from .counting import count

diff --git a/bwsample/ranking.py b/bwsample/ranking.py
@@ -323,6 +323,8 @@ def maximize_hoaglinapprox(cnt: scipy.sparse.csr_matrix,
     P.data = np.maximum(0.0, np.minimum(1.0, P.data))
     # only if Nij>Nji
     P = P.multiply(cnt > cnt.T)
+    # Q = 1-P
+    P.data = 1. - P.data
 
     # sum rows in DoK matrix
     metrics = np.array(P.sum(axis=1).flatten())[0]

diff --git a/docs/rank.ipynb b/docs/rank.ipynb
@@ -106,21 +106,114 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "  Positions: [0 5 3 1 2 4]\n",
-      "Ordered IDs: ['A' 'F' 'D' 'B' 'C' 'E']\n",
-      "     Scores: [1.  0.8 0.6 0.3 0.3 0. ]\n"
+      "  Positions: [5 3 1 2 0 4]\n",
+      "Ordered IDs: ['F' 'D' 'B' 'C' 'A' 'E']\n",
+      "     Scores: [1.  0.8 0.4 0.4 0.2 0. ]\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "../bwsample/ranking.py:72: UserWarning: Use 'approx' because it's faster.\n",
+      "  warnings.warn(\"Use 'approx' because it's faster.\", UserWarning)\n"
      ]
     }
    ],
    "source": [
     "positions, sortedids, metrics, scores, info = bws.rank(\n",
-    "    dok, method='approx', avg='exist', adjust='quantile')\n",
+    "    dok, method='pvalue', avg='exist', adjust='quantile')\n",
     "\n",
     "print(f\"  Positions: {positions}\") \n",
     "print(f\"Ordered IDs: {sortedids}\") \n",
     "print(f\"     Scores: {scores}\") "
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "7094d503-a956-45a4-9bf8-65d8dae4b070",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[0.   , 0.   , 0.   , 0.398, 0.993, 0.   ],\n",
+       "       [0.892, 0.   , 0.   , 0.   , 0.954, 0.   ],\n",
+       "       [0.892, 0.   , 0.   , 0.   , 0.954, 0.   ],\n",
+       "       [0.   , 0.892, 0.892, 0.   , 0.999, 0.   ],\n",
+       "       [0.   , 0.   , 0.   , 0.   , 0.   , 0.   ],\n",
+       "       [0.999, 0.954, 0.954, 0.993, 1.   , 0.   ]])"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "info['P'].todense().round(3)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9fdbe524-f05a-4a2d-bcd1-b06b1c071c47",
+   "metadata": {},
+   "source": [
+    "# Hoaglin Approximation"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "da48ee5c-cef2-42bc-865f-581ab47d2a1a",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "  Positions: [5 3 1 2 0 4]\n",
+      "Ordered IDs: ['F' 'D' 'B' 'C' 'A' 'E']\n",
+      "     Scores: [1.  0.8 0.4 0.4 0.2 0. ]\n"
+     ]
+    }
+   ],
+   "source": [
+    "positions, sortedids, metrics, scores, info = bws.rank(\n",
+    "    dok, method='hoaglin', avg='exist', adjust='quantile')\n",
+    "\n",
+    "print(f\"  Positions: {positions}\") \n",
+    "print(f\"Ordered IDs: {sortedids}\") \n",
+    "print(f\"     Scores: {scores}\") "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "7d68d4bc-7172-440c-95fb-6aebcb317666",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[0.   , 0.   , 0.   , 0.848, 0.971, 0.   ],\n",
+       "       [0.933, 0.   , 0.   , 0.   , 0.951, 0.   ],\n",
+       "       [0.933, 0.   , 0.   , 0.   , 0.951, 0.   ],\n",
+       "       [0.   , 0.933, 0.933, 0.   , 0.981, 0.   ],\n",
+       "       [0.   , 0.   , 0.   , 0.   , 0.   , 0.   ],\n",
+       "       [0.981, 0.951, 0.951, 0.971, 0.992, 0.   ]])"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "info['P'].todense().round(3)"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "ordered-worth",
@@ -131,7 +224,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 9,
    "id": "several-safety",
    "metadata": {},
    "outputs": [
@@ -156,7 +249,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 10,
    "id": "handy-regard",
    "metadata": {},
    "outputs": [
@@ -183,7 +276,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 11,
    "id": "convenient-hurricane",
    "metadata": {},
    "outputs": [
@@ -208,7 +301,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 12,
    "id": "gross-fiction",
    "metadata": {},
    "outputs": [
@@ -234,7 +327,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 13,
    "id": "desirable-company",
    "metadata": {},
    "outputs": [
@@ -259,7 +352,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 14,
    "id": "available-spiritual",
    "metadata": {},
    "outputs": [
@@ -277,7 +370,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 15,
    "id": "healthy-faculty",
    "metadata": {},
    "outputs": [
@@ -317,7 +410,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.3"
+   "version": "3.7.9"
   }
  },
  "nbformat": 4,