Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

finished questions #3

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

finished questions #3

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
123 changes: 79 additions & 44 deletions practise/KNN+Practise.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,14 +24,13 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:00.462641Z",
"start_time": "2017-03-09T12:11:00.457060-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
"from __future__ import print_function\n",
"import os\n",
"data_path = ['data']"
"data_path = ['../data']"
]
},
{
Expand All @@ -53,8 +52,7 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:00.913456Z",
"start_time": "2017-03-09T12:11:00.883452-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
Expand Down Expand Up @@ -86,8 +84,7 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:01.564122Z",
"start_time": "2017-03-09T12:11:01.557967-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
Expand All @@ -109,6 +106,14 @@
"data.columns"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Due to things like different income and competetion, customers in some states/areas might be more prone to terminate their service with similer tarrifs than in other areas. So the state and area code can be useful information in predicting whether a customer will terminate the service or not.\n",
"Phone number will obviously not be a good feature to analyze since it won't have an affect on if a customer leaves and will only serve to pollute the data set. "
]
},
{
"cell_type": "markdown",
"metadata": {},
Expand All @@ -126,17 +131,21 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:04.545751Z",
"start_time": "2017-03-09T12:11:04.509105-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
"from sklearn.preprocessing import LabelBinarizer\n",
"\n",
"lb = LabelBinarizer()\n",
"\n",
"for col in ['intl_plan', 'voice_mail_plan', 'churned']:\n",
" data[col] = lb.fit_transform(data[col])"
" print(data.head(2)[col])\n",
"print()\n",
"for col in ['intl_plan', 'voice_mail_plan', 'churned']:\n",
" data[col] = lb.fit_transform(data[col])\n",
"print()\n",
"for col in ['intl_plan', 'voice_mail_plan', 'churned']:\n",
" print(data.head(2)[col])"
]
},
{
Expand All @@ -146,8 +155,7 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:04.736451Z",
"start_time": "2017-03-09T12:11:04.718049-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
Expand All @@ -160,7 +168,8 @@
"msc = MinMaxScaler()\n",
"\n",
"data = pd.DataFrame(msc.fit_transform(data), # this is an np.array, not a dataframe.\n",
" columns=data.columns)"
" columns=data.columns)\n",
"print(data.head(2))"
]
},
{
Expand All @@ -180,8 +189,7 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:50.280188Z",
"start_time": "2017-03-09T12:11:50.269326-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
Expand All @@ -191,7 +199,6 @@
"# Split the data into two dataframes\n",
"X_data = data[x_cols]\n",
"y_data = data['churned']\n",
"\n",
"# # alternatively:\n",
"# X_data = data.copy()\n",
"# y_data = X_data.pop('churned')"
Expand All @@ -204,8 +211,7 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:50.989446Z",
"start_time": "2017-03-09T12:11:50.498708-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
Expand Down Expand Up @@ -237,15 +243,14 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:50.997204Z",
"start_time": "2017-03-09T12:11:50.991392-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
"# Function to calculate the % of values that were correctly predicted\n",
"\n",
"def accuracy(real, predict):\n",
" return sum(y_data == y_pred) / float(real.shape[0])"
" return sum(real == predict) / float(real.shape[0])"
]
},
{
Expand Down Expand Up @@ -281,12 +286,17 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:52.047123Z",
"start_time": "2017-03-09T12:11:51.538212-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
"#Student writes code here"
"#Student writes code here\n",
"knn = KNeighborsClassifier(n_neighbors=3, weights='distance')\n",
"\n",
"knn = knn.fit(X_data, y_data)\n",
"\n",
"y_pred = knn.predict(X_data)\n",
"print(accuracy(y_data, y_pred))"
]
},
{
Expand All @@ -296,11 +306,18 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:11:52.755941Z",
"start_time": "2017-03-09T12:11:52.049816-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": []
"source": [
"knn = KNeighborsClassifier(n_neighbors=3, weights='distance', p=1)\n",
"\n",
"knn = knn.fit(X_data, y_data)\n",
"\n",
"y_pred = knn.predict(X_data)\n",
"\n",
"print(accuracy(y_data, y_pred))"
]
},
{
"cell_type": "markdown",
Expand All @@ -319,12 +336,17 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:12:01.329053Z",
"start_time": "2017-03-09T12:11:52.759302-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": [
"#Student writes code here"
"#Student writes code here\n",
"ksAccuracy = {} #to store the accuracy for each k\n",
"for k in range(1, 21):\n",
" knn = KNeighborsClassifier(n_neighbors=k)\n",
" knn = knn.fit(X_data, y_data)\n",
" y_pred = knn.predict(X_data)\n",
" ksAccuracy[k] = accuracy(y_data, y_pred) "
]
},
{
Expand All @@ -334,11 +356,16 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:12:01.829160Z",
"start_time": "2017-03-09T12:12:01.331021-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": []
"source": [
"#To view the table\n",
"print(\"{:<8} {:<15}\".format(\"k\", \"accuracy\"))\n",
"print()\n",
"for k, acc in ksAccuracy.items():\n",
" print(\"{:<8} {:<15}\".format(k, acc))"
]
},
{
"cell_type": "code",
Expand All @@ -347,28 +374,36 @@
"ExecuteTime": {
"end_time": "2017-03-09T17:12:02.238935Z",
"start_time": "2017-03-09T12:12:01.831094-05:00"
},
"collapsed": true
}
},
"outputs": [],
"source": []
"source": [
"#Requires matplotlib\n",
"\n",
"import matplotlib.pyplot as plt\n",
"plt.plot(ksAccuracy.keys(), ksAccuracy.values(), 'ro')\n",
"plt.xticks(range(1,21))\n",
"plt.xlabel(\"Ks\")\n",
"plt.ylabel(\"Accuracy\")\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"outputs": [],
"source": []
"source": [
"Since the \"training set\" is the same is the test set, at K=1, the nearest neighbour matches exactly with the point in the data set for every point in the set. Therefore the accuracy is 100%."
]
}
],
"metadata": {
"anaconda-cloud": {},
"kernelspec": {
"display_name": "Python 3",
"display_name": "Python 2",
"language": "python",
"name": "python3"
"name": "python2"
},
"language_info": {
"codemirror_mode": {
Expand All @@ -380,7 +415,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.3"
"version": "3.5.2"
},
"name": "Linear_Regression_and_K_Nearest_Neighbors_Exercises-ANSWERS",
"notebookId": 2125319687183902
Expand Down