Skip to content

Instantly share code, notes, and snippets.

@ljbelenky
Created March 24, 2021 22:15
Show Gist options
  • Save ljbelenky/e1bf5eeb4a24bf53ec9e7380af8f0a39 to your computer and use it in GitHub Desktop.
Save ljbelenky/e1bf5eeb4a24bf53ec9e7380af8f0a39 to your computer and use it in GitHub Desktop.
Using Unsupervised Learning to Inform Supervised Learning
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Using Clustering to Inform Supervised Learning\n",
"\n",
"## Problem Statement:\n",
"\n",
"We want to build a regression model that shows expected production from an oil well **during normal production**\n",
"\n",
"### Challenge:\n",
"\n",
"We have good data to show oil well production, but this does not just show when the oil well is working. It also includes data from periods when the well is down (not producing), producing less than expectd, or producing more than expected.\n",
"\n",
"To model periods of normal production, we must first identify \"normal\" production, and fit a model only to these data points. \n",
"\n",
"Unfortunately, we do not have a label to indicate when the well is in normal production. Therefore, to to generate these labels, we need to use unsupervised learning."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Generate Some Data:\n",
"\n",
"Let's generate some synthetic data:"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd \n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"from sklearn.decomposition import PCA\n",
"from sklearn.cluster import KMeans\n",
"from sklearn.linear_model import LinearRegression as LR\n",
"np.random.seed(42)\n",
"\n",
"class Decline_Generator():\n",
" '''This class creates synthetic gas production data with dropouts, low point, high points and noise.\n",
" Return data is a pandas dataframe with columns \"date\" and \"production'''\n",
" def __init__(self, n = 365*2, drops = (0,0), \n",
" noise = 0, highpoints = (0,0), lowpoints = (0,0), decline_rate = .996):\n",
" dates = pd.date_range(start = '2018-01-01', end = '2019-01-01', periods = n)\n",
" \n",
" x = 100\n",
" production = []\n",
" \n",
" for _ in range(n):\n",
" production.append(x)\n",
" x*=decline_rate\n",
" \n",
" production = np.array(production) + np.random.random(size = n)*noise\n",
"\n",
" for i in range(highpoints[0]):\n",
" index = np.random.randint(n)\n",
" production[index] += highpoints[1]\n",
"\n",
" for i in range(lowpoints[0]):\n",
" index = np.random.randint(n)\n",
" production[index] -= lowpoints[1]\n",
"\n",
" for i in range(drops[0]):\n",
" start = np.random.randint(n-drops[1])\n",
" production[start:start+drops[1]] = 0\n",
" \n",
" production += np.random.random(size = n)*noise\n",
" production[production <=1] = 1.1\n",
" \n",
" data = pd.DataFrame({'date':dates,'production':production})\n",
" data['date'] = pd.to_datetime(data['date'].dt.date)\n",
" \n",
" \n",
" self.data = data "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## An ideal oil well:\n",
"\n",
"Ideally, an oil well would show predictable, normal production:"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"data = Decline_Generator().data\n",
"\n",
"plt.plot(data['date'],data['production'])\n",
"plt.ylabel('Production') \n",
"plt.title('Idealized Oil Well Production Curve');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Real Oil Well Production\n",
"\n",
"A real oil well is likely to have periods of high, normal, low and zero production:\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"data = Decline_Generator(drops = (15,25), noise = 10, highpoints = (35,20), lowpoints = (25,20)).data\n",
"\n",
"plt.plot(data['date'],data['production'])\n",
"plt.ylabel('Production') \n",
"plt.title('Realistic Oil Well Production Curve');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Simple Fitting\n",
"\n",
"If we fit a simple linear model to all the data, it doesn't work well because it fits to both the \"normal\" production and the abnormal production"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"from sklearn.linear_model import LinearRegression as LR\n",
"linear = LR().fit(data[['date']].astype(int),data['production'] )\n",
"plt.plot(data['date'], linear.predict(data[['date']].astype(int)))\n",
"plt.plot(data['date'], data['production'])\n",
"plt.title('A Pretty Dumb, Linear Fit');"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"exp_model = LR().fit(data[['date']].astype(int),np.log(data['production'] ))\n",
"plt.plot(data['date'], np.exp(exp_model.predict(data[['date']].astype(int))))\n",
"plt.plot(data['date'], data['production'])\n",
"plt.title('A Pretty Dumb, Exponential Fit');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Non-Linear Regression:\n",
"\n",
"Using a non-linear Regression doesn't help. It does a better job of fitting all the data, but is not helpful for identifying ideal production. I.e.: when the model production drops to zero, the prediction line also drops to zero. We want that line to show where the production *should be*, not where it is."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"from sklearn.ensemble import RandomForestRegressor as RFR\n",
"rfr = RFR(n_estimators = 10, max_depth=10).fit(data[['date']].astype(int), data['production'] )\n",
"plt.plot(data['date'], rfr.predict(data[['date']].astype(int)))\n",
"plt.plot(data['date'], data['production'], marker = '.', linestyle = '')\n",
"plt.title('A Pretty Dumb, Random Forest');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Solution:\n",
"\n",
"To solve this problem, we will use **unsupervised learning** to identify different types of production, then fit a regression model *only to the normal data points*.\n",
"\n",
"First, we have to engineer some additional features:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"X = data.copy()\n",
"\n",
"X['production'] = np.log(X['production'])\n",
"\n",
"'''Add Delta Columns for differences''' \n",
"for i in range(-4,5):\n",
" X['delta{}'.format(i)] = X.production.diff(i)\n",
"\n",
"X['days_ago'] = (X['date'].max()-X['date']).dt.days\n",
"X['epsilon'] = X['days_ago']/X['production']\n",
"\n",
"X = X.drop(['date','days_ago'], axis = 1).reset_index()\n",
"X = X.dropna()\n",
"indices = X['index']\n",
"X.drop('index',axis = 1, inplace = True)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>production</th>\n",
" <th>delta-4</th>\n",
" <th>delta-3</th>\n",
" <th>delta-2</th>\n",
" <th>delta-1</th>\n",
" <th>delta0</th>\n",
" <th>delta1</th>\n",
" <th>delta2</th>\n",
" <th>delta3</th>\n",
" <th>delta4</th>\n",
" <th>epsilon</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>4.708870</td>\n",
" <td>0.001104</td>\n",
" <td>0.020480</td>\n",
" <td>-0.012120</td>\n",
" <td>0.096722</td>\n",
" <td>0.0</td>\n",
" <td>-0.009374</td>\n",
" <td>0.045422</td>\n",
" <td>-0.027363</td>\n",
" <td>0.039633</td>\n",
" <td>77.088552</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>4.612149</td>\n",
" <td>-0.038065</td>\n",
" <td>-0.095617</td>\n",
" <td>-0.076242</td>\n",
" <td>-0.108841</td>\n",
" <td>0.0</td>\n",
" <td>-0.096722</td>\n",
" <td>-0.106095</td>\n",
" <td>-0.051300</td>\n",
" <td>-0.124084</td>\n",
" <td>78.705178</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>4.720990</td>\n",
" <td>0.016636</td>\n",
" <td>0.070776</td>\n",
" <td>0.013224</td>\n",
" <td>0.032600</td>\n",
" <td>0.0</td>\n",
" <td>0.108841</td>\n",
" <td>0.012120</td>\n",
" <td>0.002746</td>\n",
" <td>0.057541</td>\n",
" <td>76.678831</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>4.688390</td>\n",
" <td>0.046712</td>\n",
" <td>-0.015963</td>\n",
" <td>0.038176</td>\n",
" <td>-0.019376</td>\n",
" <td>0.0</td>\n",
" <td>-0.032600</td>\n",
" <td>0.076242</td>\n",
" <td>-0.020480</td>\n",
" <td>-0.029854</td>\n",
" <td>77.212000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>4.707766</td>\n",
" <td>-0.025336</td>\n",
" <td>0.066088</td>\n",
" <td>0.003412</td>\n",
" <td>0.057552</td>\n",
" <td>0.0</td>\n",
" <td>0.019376</td>\n",
" <td>-0.013224</td>\n",
" <td>0.095617</td>\n",
" <td>-0.001104</td>\n",
" <td>76.681802</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" production delta-4 delta-3 delta-2 delta-1 delta0 delta1 \\\n",
"4 4.708870 0.001104 0.020480 -0.012120 0.096722 0.0 -0.009374 \n",
"5 4.612149 -0.038065 -0.095617 -0.076242 -0.108841 0.0 -0.096722 \n",
"6 4.720990 0.016636 0.070776 0.013224 0.032600 0.0 0.108841 \n",
"7 4.688390 0.046712 -0.015963 0.038176 -0.019376 0.0 -0.032600 \n",
"8 4.707766 -0.025336 0.066088 0.003412 0.057552 0.0 0.019376 \n",
"\n",
" delta2 delta3 delta4 epsilon \n",
"4 0.045422 -0.027363 0.039633 77.088552 \n",
"5 -0.106095 -0.051300 -0.124084 78.705178 \n",
"6 0.012120 0.002746 0.057541 76.678831 \n",
"7 0.076242 -0.020480 -0.029854 77.212000 \n",
"8 -0.013224 0.095617 -0.001104 76.681802 "
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"X.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cluster the Data Points into 6 Clusters"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"X = PCA(n_components = int(len(X.columns)/1.5), whiten = True).fit_transform(X)\n",
"\n",
"labels = KMeans(n_clusters =6).fit(X).predict(X)\n",
"data['labels'] = -1\n",
"data.iloc[indices,-1] = labels\n",
"data = data[data['labels']>=0]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Assume the majority class is \"normal production\""
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"majority_class = data['labels'].value_counts().index[0]\n",
"majority = data[data['labels']==majority_class]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Fit a model only on the normal production data"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"model = LR().fit(majority['date'].values.reshape(-1,1), np.log(majority['production']))\n",
"predictions = np.exp(model.predict(majority['date'].values.astype(float).reshape(-1,1)))\n",
"\n",
"for label in data['labels'].unique():\n",
" l = data[data['labels'] == label]\n",
" plt.scatter(l['date'], l['production'], marker = '.')\n",
"\n",
"plt.plot(majority['date'], predictions, linewidth = 3, color = 'r');\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Conclusion:\n",
"\n",
"We have sucessfully fit a model to the normal production data. "
]
}
],
"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": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment