TD4_Sklearn_suite.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "\n",
    "class AddFeatures(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, where_x, functions=None):\n",
    "        self.functions = functions\n",
    "        self.where_x = where_x\n",
    "    def fit(self, X, y):\n",
    "        return self\n",
    "    def transform(self, X):\n",
    "        # .T to iterate over the cols of the array\n",
    "        # vstack to stack vertically row by row\n",
    "        # then .T to go back to the initial shape\n",
    "        # not taking the intercept (index 0)\n",
    "        #if self.intercept_exist:\n",
    "        #    start = 1\n",
    "        #    x_ = X[:,start:]\n",
    "        #else:\n",
    "        #    start = 0\n",
    "        #    x_ = X\n",
    "        x_ = X[:, [self.where_x]].copy()\n",
    "        self.funcs_applied = [\"{}\".format(func.__name__) for func in self.functions]\n",
    "        cols =  np.hstack([func(x_) for func in self.functions])\n",
    "        return np.hstack([X, cols])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class BetterPipeline(Pipeline):\n",
    "    \"\"\" Since pipeline.transform does work only when all estimators are transformers of the data,\n",
    "    and mostly i'm fitting a model as last estimator, i prefer creating a just_transform method for that\"\"\"\n",
    "    def just_transforms(self, X):\n",
    "        \"\"\"Applies all transforms to the data, without applying last \n",
    "           estimator.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        X : iterable\n",
    "            Data to predict on. Must fulfill input requirements of first step of\n",
    "            the pipeline.\n",
    "        \"\"\"\n",
    "        Xt = X\n",
    "        for name, transform in self.steps[:-1]:\n",
    "            Xt = transform.transform(Xt)\n",
    "        return Xt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.linear_model import SGDRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.linspace(0, 2*np.pi, 100)\n",
    "y = np.random.normal(np.sin(x), scale=0.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "funcs = [np.sin, np.cos, np.exp]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_after_feature_engineering = pipeline.just_transforms(x[:, np.newaxis])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'pipeline' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-17-2ef177e65378>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m \u001b[0mpipeline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_params\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0madding_features__functions\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfuncs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m \u001b[0mpipeline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnewaxis\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'pipeline' is not defined"
     ]
    }
   ],
   "source": [
    "# scenario 1\n",
    "scaling_and_gradient_descent = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('scaler', StandardScaler()),\n",
    "    ('linear_reg', SGDRegressor(fit_intercept=False))\n",
    "])\n",
    "\n",
    "# scenario 2\n",
    "scaling_and_OLS = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('scaler', StandardScaler()),\n",
    "    ('linear_reg', SGDRegressor(fit_intercept=False))\n",
    "])\n",
    "\n",
    "# scenario 3\n",
    "OLS = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('linear_reg', LinearRegression(fit_intercept=False))\n",
    "])\n",
    "\n",
    "# scenario 3\n",
    "gradient_descent = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('linear_reg', LinearRegression(fit_intercept=False))\n",
    "])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "pipeline.set_params(adding_features__functions=funcs)\n",
    "pipeline.fit(x[:, np.newaxis], y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tuning hyperparameters or data processing steps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ok so, so far we fitted a **model on a train** set and later computed an **MSE** on both the train, and **test** set to assess whether an **overfit** would have occured. And that was the case. \n",
    "\n",
    "So we decided to use another model than the base linear regression one. We used **Lasso**.<br>\n",
    "Lasso does have a **slightly difference risk function definition**: to the sum of squared errors risk function derived from OLS, Lasso **adds an additional penalty to constraint the amplitude (1-norm) of the estimated $\\beta$s** while still trying to meet **minimization** of the sum of squared errors.\n",
    "\n",
    "OLS estimator, being the **best linear unbiased estimator**, **adding an additional constraint** for the $\\beta$s will **add bias to those estimations** of the $\\beta$s (what we call **estimation bias**) which **adds up to** the bias the model may already have (**from assumptions, ommited variables or interactions** or other), what we call **model bias** (the difference from our best-fitting linear approximation and the true function).\n",
    "\n",
    "Why doing so ? In an attempt to **decrease variances around the estimates**, hence the **variance of the model itself**, and get a **lower MSE eventually**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The penality coefficient $lambda$ (or $alpha$ depending on the litterature or the API) is what we call an **hyperparameter**: we fix it **prior** to the **actual learning process**.<br> Here this hyperparameter is a called a regularization hyperparameter.<br>\n",
    "In Scikit-Learn, they are passed as arguments to the constructor of the estimator classes.<br>\n",
    "So **which** $lambda$ to use so to get the most reduction in a MSE compared to our initial linear regression model ? <br>\n",
    "and by the **way which MSE** are we talking about ? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Of course we are not going to control setting up our hyperparameters based on the MSE for the training set: that would lead exactly to the very first situation where we **shaped our mind, our representation of our data and our modelisation out of it to solely satisfy ourself on what we know, rather than what we don't yet** loosing all the predictive ability of our model, and getting back to the overfitting issue."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cross validation to the rescourse !"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cross-validation is the simplest method for estimating the **expected prediction error** i.e. the expected extra-sample error for when we fit a model to a training set and evaluate its predictions on **unseen data**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So we will check the MSE on the test set, in a bid to reduce it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But if you were to **tune** hyperparams or data preparation steps while **checking variations of MSE on test towards a minimization of it**, well, we would still somehow use a metric, a quantitative measure **we shouldn't be aware of**, as it is supposed to be the mean squared errors of the model on **unseen data**.<br>\n",
    "To give another example: it is as if you had to forecast whether or not to buy vegetables while not having access to the inside of the fridge. If you can **weight** the fridge itself, you might not know how many vegetables are left among all the food, but at least you have a taste of how likely the fridge is empty, considering the vegetables are the heaviest, hence you modify your behavior respectively.\n",
    "\n",
    "This has a name: it is called **data leakage**.\n",
    "\n",
    "You would have to actually split the whole data in 3 sets: **train**, **test** and **validation**, so to keep at least one set of data only for estimating the expected prediction error of the final model.<br>\n",
    "You train the model with $lambda1$ on the training set, you monitor the MSE on the test set, you update $lambda$ to the new value, and once you found a satisfying minimum of the MSE, you can retrain on the whole available data (train+test) and finally evaluate the final model using the hold-out validation test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 969,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.linear_model import SGDRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.linear_model import Lasso"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "funcs = [np.sin, np.cos, np.exp]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [],
   "source": [
    "# scenario 1\n",
    "scaling_and_gradient_descent = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0, functions=funcs)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('scaler', StandardScaler()),\n",
    "    ('linear_reg', SGDRegressor(fit_intercept=False))\n",
    "])\n",
    "\n",
    "# scenario 2\n",
    "scaling_and_OLS = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0, functions=funcs)),\n",
    "    ('poly', PolynomialFeatures(degree=2, interaction_only=True)),\n",
    "    ('scaler', StandardScaler()),\n",
    "    ('linear_reg', LinearRegression(fit_intercept=False))\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BetterPipeline(steps=[('adding_features',\n",
       "                       AddFeatures(functions=[<ufunc 'sin'>, <ufunc 'cos'>,\n",
       "                                              <ufunc 'exp'>],\n",
       "                                   where_x=0)),\n",
       "                      ('poly', PolynomialFeatures(interaction_only=True)),\n",
       "                      ('scaler', StandardScaler()),\n",
       "                      ('linear_reg', LinearRegression(fit_intercept=False))])"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaling_and_gradient_descent.fit(x2, y)\n",
    "scaling_and_OLS.fit(x2, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_transformed = scaling_and_gradient_descent.just_transforms(x2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### pour Denis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Lasso(alpha=1e-08, tol=1)"
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lasso = Lasso(alpha=0.00000001, tol=1)\n",
    "lasso.fit(x_transformed, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0.5, 1.0, 'coefficients obtained using Gradient Descent')]"
      ]
     },
     "execution_count": 328,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=1, ncols=3, figsize=(15,3), sharey=True)\n",
    "sns.barplot(x=list(range(11)), y=lasso.coef_, ax=axes[0]).set(\n",
    "    title=\"coefficients obtained using Lasso, lambda 1e-18\")\n",
    "sns.barplot(x=list(range(11)), y=scaling_and_gradient_descent.named_steps.linear_reg.coef_, \n",
    "            ax=axes[1]).set(title=\"coefficients obtained using Gradient Descent\")\n",
    "#sns.barplot(x=list(range(11)), y=scaling_and_OLS.named_steps.linear_reg.coef_, \n",
    "#            ax=axes[2]).set(title=\"coefficients obtained using OLS\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(x_transformed, y, train_size=0.70)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 392,
   "metadata": {},
   "outputs": [],
   "source": [
    "lasso_models, alphas, MSE_train, MSE_test = [], [], [], []\n",
    "for alpha in np.linspace(0.0000001,0.25,100):\n",
    "    lasso_model = Lasso(alpha=alpha, tol=0.5)\n",
    "    lasso_model.fit(X_train, y_train)\n",
    "    mse_train = mean_squared_error( lasso_model.predict(X_train), y_train )\n",
    "    mse_test  = mean_squared_error( lasso_model.predict(X_test), y_test )\n",
    "    alphas.append(alpha); MSE_train.append(mse_train); MSE_test.append(mse_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 419,
   "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": [
    "plt.plot(alphas, MSE_train, color='r', label=\"MSE train\")\n",
    "plt.plot(alphas, MSE_test,  color='b', label=\"MSE test\")\n",
    "plt.ylim(0.10, 0.35)\n",
    "plt.vlines(x=np.array(alphas)[np.argsort(MSE_test)[:20:2]], ymin=0, ymax=min(MSE_test), color='b', linestyles='-.', label=\"minimal values of MSE test\")\n",
    "plt.legend()\n",
    "plt.suptitle(\"MSE on train and test sets\", fontsize=14)\n",
    "plt.title(\"For some of Lasso regularization hyperparameter\")\n",
    "plt.tight_layout(pad=0.6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# K-Fold cross validation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Splitting data again and again in an attempt to put Chinese walls in your ML workflow, lead to another issue: what if you **don't have much** data? it is likely your MSE(test) on a **few dozen points could be overly optimistic**, what if by chance you got the right test points to have a sweet MSE that suit your needs for a certain model ? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K-Fold cross validation is an attempt to use every data points at least in the testing part.<br>\n",
    "It is still cross-validation, but this time you split your dataset in **K** aproximately equally sized **folds**.<br>\n",
    "Then you **train the model on K-1** folds and test it on the **remaining one**. You do it **K times** (each time tested for **each remaining test fold**). Yes, you end up with **K number of MSE(test-fold)**.\n",
    "- On a train-test design, the average of the MSE is still a good estimate for the expected prediction error of the model.\n",
    "- On a train-test-validation one, you still have more confidence that no data points were left while computing the MSE \"on the test\"."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's do a k-fold cross validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score, cross_val_predict\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.metrics import mean_squared_error, make_scorer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.95596911, -2.89048347,  0.18178379, -5.47694298, -2.31706554])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(LinearRegression(), x2, y, scoring=\"r2\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "scoring takes a scoring parameter (greater is better), hence is used the R2 is an appropriate choice, \n",
    "we could have taken the negation of the MSE too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.3279406 , 0.45555535, 0.25739018, 1.03002024, 1.27365498])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(LinearRegression(), x2, y, scoring=make_scorer(mean_squared_error))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.3279406 , 0.45555535, 0.25739018, 1.03002024, 1.27365498])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "- cross_val_score(LinearRegression(), x2, y, scoring=\"neg_mean_squared_error\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 5 folds\n",
    "- 5 model training\n",
    "- 5 test sets\n",
    "- 5 MSE\n",
    "\n",
    "Wow ? difference are so important from one test set to another ! why is so ?\n",
    "When the MSE is high, the R2 is low, sometimes negative ? worse than a simple dummy model (H0 hypthesis) using the average. Why is so ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "folding = KFold(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<generator object _BaseKFold.split at 0x12b2c5430>"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "folding.split(x2, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Wow ! a generator object !"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator = folding.split(x2, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36,\n",
       "        37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,\n",
       "        54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70,\n",
       "        71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87,\n",
       "        88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99]),\n",
       " array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
       "        17, 18, 19]))"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "next(generator) # training indices, testing indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_val_visualize(X, y, cv=5, shuffle=False):\n",
    "    from sklearn.model_selection import KFold\n",
    "    from sklearn.metrics import mean_squared_error\n",
    "    from sklearn.linear_model import LinearRegression\n",
    "\n",
    "    new_fig, axes =  plt.subplots(figsize=(cv*3, 3), ncols=cv, sharey=True)\n",
    "    regressions, MSE = [], []\n",
    "    folds = KFold(cv, shuffle=shuffle).split(X, y)\n",
    "    \n",
    "    for i_, (train_indices, test_indices) in enumerate(folds):\n",
    "        # train on each 4 folds subset\n",
    "        lm = LinearRegression()\n",
    "        lm.fit( X[train_indices], y[train_indices])\n",
    "        # predict on each test fold\n",
    "        y_pred = lm.predict(X[test_indices])\n",
    "        # compute MSE for those test fold\n",
    "        mse = mean_squared_error(lm.predict(X[test_indices]), y[test_indices])\n",
    "        # plot the training points, test points, and the fit for the fold\n",
    "        axes[i_].scatter( X[train_indices], y[train_indices], color='r')\n",
    "        axes[i_].scatter( X[test_indices], y[test_indices], color='g')\n",
    "        axes[i_].plot( X, lm.predict(X), color='black' )\n",
    "        # save the results | save the models\n",
    "        MSE.append(mse); regressions.append(lm)\n",
    "    return MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.32794060000061515,\n",
       " 0.45555535076735226,\n",
       " 0.2573901788986118,\n",
       " 1.0300202445239781,\n",
       " 1.273654982470511]"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cross_val_visualize(x2, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.24514477688923542, 0.38705639490105015, 0.6231891770127047]"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cross_val_visualize(x2, y, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.3711358856524268,\n",
       " 0.3821283100103246,\n",
       " 0.5107607283792079,\n",
       " 0.4372335783422797,\n",
       " 0.37844005566106576]"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cross_val_visualize(x2, y, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Much more homogeneous results !"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yes, the misleading results here are drawn by the fact the train and test set were not taken randomly"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bringing that up together: GridSearch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you followed the previous steps, here is what is going to be the overall scheme of the hyperparameter tuner (you :p)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Splitting the whole dataset in **train - validation** (e.g. of ratios: 0.80 / 0.20).\n",
    "2. Leave the validation set for a while, it will be use **at the end** for an estimation of the expected prediction error of the **final model**.\n",
    "3. Choose a **set of hyperparameters values** to try, if you have 2 hyperparameters $hyper1$ and $hyper2$, one would do want to try all combinations of the values taken by $(hyper1, hyper2)$\n",
    "4. Split the **training set** in **train - set**, or better, split into $k$ sets of $k-fold$ partitions.\n",
    "5. Select a model (or a set of models) and a **combination of its corresponding** hyperparameter $(hyper1, hyper2)$\n",
    "6. Train it $k$ times (one for each cross validation) and average the **MSE results**.\n",
    "7. Pick up the model which performed the best, using the average MSE with the best combination of hyperparameters' values.\n",
    "8. **Retrain** it on the whole initial training set (as it is the model you elected, you no longer need to use this/these intermediates test-sets.\n",
    "9. Evaluate the performance using the held-out validation set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Actually, this is a pattern often used and sklearn provide a function for this (rather that implementing by hand and using nested for loops to find inside the hyperparameter space...)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 939,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import mean_squared_error, make_scorer\n",
    "from sklearn.linear_model import Lasso, Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 960,
   "metadata": {},
   "outputs": [],
   "source": [
    "gs = GridSearchCV(\n",
    "    estimator=Lasso(tol=0.5),\n",
    "    param_grid={ \"alpha\" : np.linspace(0.000001, 0.25, 100) },\n",
    "    scoring=make_scorer(mean_squared_error),\n",
    "    cv=KFold(5, shuffle=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 961,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=5, random_state=None, shuffle=True),\n",
       "             estimator=Lasso(tol=0.5),\n",
       "             param_grid={'alpha': array([1.00000000e-06, 2.52624242e-03, 5.05148485e-03, 7.57672727e-03,\n",
       "       1.01019697e-02, 1.26272121e-02, 1.51524545e-02, 1.76776970e-02,\n",
       "       2.02029394e-02, 2.27281818e-02, 2.52534242e-02, 2.77786667e-02,\n",
       "       3.03039091e-02, 3.28291515e-02, 3.53543939e-02, 3.787963...\n",
       "       1.91919424e-01, 1.94444667e-01, 1.96969909e-01, 1.99495152e-01,\n",
       "       2.02020394e-01, 2.04545636e-01, 2.07070879e-01, 2.09596121e-01,\n",
       "       2.12121364e-01, 2.14646606e-01, 2.17171848e-01, 2.19697091e-01,\n",
       "       2.22222333e-01, 2.24747576e-01, 2.27272818e-01, 2.29798061e-01,\n",
       "       2.32323303e-01, 2.34848545e-01, 2.37373788e-01, 2.39899030e-01,\n",
       "       2.42424273e-01, 2.44949515e-01, 2.47474758e-01, 2.50000000e-01])},\n",
       "             scoring=make_scorer(mean_squared_error))"
      ]
     },
     "execution_count": 961,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs.fit(x_transformed, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 962,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Lasso(alpha=0.25, tol=0.5), {'alpha': 0.25})"
      ]
     },
     "execution_count": 962,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs.best_estimator_, gs.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 963,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_grid = pd.DataFrame(gs.cv_results_)\n",
    "params = df_grid.params.apply(pd.Series)\n",
    "df_grid = pd.concat([params, df_grid], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 964,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lucbertin/.pyenv/versions/3.8.4/lib/python3.8/site-packages/pandas/plotting/_matplotlib/core.py:1235: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.set_xticklabels(xticklabels)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "splits_measures = [col for col in df_grid if col.startswith(\"split\")] \n",
    "ax = df_grid.set_index([\"alpha\"])[splits_measures].T.plot(kind=\"line\", figsize=(10,4))\n",
    "ax.get_legend().remove()\n",
    "#plt.xticks(rotation=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 965,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=SVR(),\n",
       "             param_grid={'C': array([ 4.        ,  4.57894737,  5.15789474,  5.73684211,  6.31578947,\n",
       "        6.89473684,  7.47368421,  8.05263158,  8.63157895,  9.21052632,\n",
       "        9.78947368, 10.36842105, 10.94736842, 11.52631579, 12.10526316,\n",
       "       12.68421053, 13.26315789, 13.84210526, 14.42105263, 15.        ]),\n",
       "                         'gamma': array([0.001     , 0.01410526, 0.02721053, 0.04031579, 0.05342105,\n",
       "       0.06652632, 0.07963158, 0.09273684, 0.10584211, 0.11894737,\n",
       "       0.13205263, 0.14515789, 0.15826316, 0.17136842, 0.18447368,\n",
       "       0.19757895, 0.21068421, 0.22378947, 0.23689474, 0.25      ])},\n",
       "             scoring=make_scorer(mean_squared_error))"
      ]
     },
     "execution_count": 965,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "param_grid = {\n",
    "    'C'     : np.linspace(4, 15, 20),\n",
    "    'gamma' : np.linspace(0.001, 0.25, 20)\n",
    "}\n",
    "gs = GridSearchCV(estimator=SVR(), param_grid=param_grid, cv=3, scoring=make_scorer(mean_squared_error))\n",
    "gs.fit(x_transformed, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 966,
   "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>C</th>\n",
       "      <th>gamma</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.611261</td>\n",
       "      <td>0.261456</td>\n",
       "      <td>0.596544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.014105</td>\n",
       "      <td>0.173154</td>\n",
       "      <td>0.232858</td>\n",
       "      <td>0.203825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.027211</td>\n",
       "      <td>0.138710</td>\n",
       "      <td>0.243590</td>\n",
       "      <td>0.192375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.040316</td>\n",
       "      <td>0.141636</td>\n",
       "      <td>0.228891</td>\n",
       "      <td>0.220583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.053421</td>\n",
       "      <td>0.144939</td>\n",
       "      <td>0.238139</td>\n",
       "      <td>0.266923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>15.0</td>\n",
       "      <td>0.197579</td>\n",
       "      <td>0.404329</td>\n",
       "      <td>0.256850</td>\n",
       "      <td>0.741546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>396</th>\n",
       "      <td>15.0</td>\n",
       "      <td>0.210684</td>\n",
       "      <td>0.422260</td>\n",
       "      <td>0.257666</td>\n",
       "      <td>0.784791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>397</th>\n",
       "      <td>15.0</td>\n",
       "      <td>0.223789</td>\n",
       "      <td>0.441236</td>\n",
       "      <td>0.257839</td>\n",
       "      <td>0.856519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>398</th>\n",
       "      <td>15.0</td>\n",
       "      <td>0.236895</td>\n",
       "      <td>0.461408</td>\n",
       "      <td>0.259146</td>\n",
       "      <td>0.911760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>15.0</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.485542</td>\n",
       "      <td>0.261651</td>\n",
       "      <td>0.934773</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>400 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        C     gamma  split0_test_score  split1_test_score  split2_test_score\n",
       "0     4.0  0.001000           0.611261           0.261456           0.596544\n",
       "1     4.0  0.014105           0.173154           0.232858           0.203825\n",
       "2     4.0  0.027211           0.138710           0.243590           0.192375\n",
       "3     4.0  0.040316           0.141636           0.228891           0.220583\n",
       "4     4.0  0.053421           0.144939           0.238139           0.266923\n",
       "..    ...       ...                ...                ...                ...\n",
       "395  15.0  0.197579           0.404329           0.256850           0.741546\n",
       "396  15.0  0.210684           0.422260           0.257666           0.784791\n",
       "397  15.0  0.223789           0.441236           0.257839           0.856519\n",
       "398  15.0  0.236895           0.461408           0.259146           0.911760\n",
       "399  15.0  0.250000           0.485542           0.261651           0.934773\n",
       "\n",
       "[400 rows x 5 columns]"
      ]
     },
     "execution_count": 966,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_grid = show_params_as_df(gs, ['C', 'gamma'])\n",
    "df_grid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 967,
   "metadata": {},
   "outputs": [],
   "source": [
    "pivot = df_grid.pivot_table(index='C', columns='gamma').stack(level=1).apply(np.mean, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 968,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='gamma', ylabel='C'>"
      ]
     },
     "execution_count": 968,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(pivot.unstack())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Enhanced GridSearch over pipeline params ! :O"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> Parameters of the estimators in the pipeline can be accessed using the **estimator__parameter** syntax. **Individual steps may also be replaced** as parameters, and **non-final steps may be ignored** by setting them to **'passthrough'**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's take 2 scenarios set (each defined as a pipeline), the possibilities are endless by combining GridSearch with Pipelines:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 925,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using GridSearch\n",
    "pipeline = BetterPipeline([\n",
    "    ('adding_features', AddFeatures(where_x=0)),\n",
    "    ('poly', PolynomialFeatures()),\n",
    "    ('scaler', StandardScaler()),\n",
    "    ('linear_reg', LinearRegression())\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 926,
   "metadata": {},
   "outputs": [],
   "source": [
    "def identity(x):\n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 927,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid = dict(\n",
    "    adding_features__functions = [[identity], [identity, np.sin, np.exp, np.cos], [np.sin], [np.exp]],\n",
    "    poly__degree = [1,2,3]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 928,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=3, random_state=None, shuffle=True),\n",
       "             estimator=BetterPipeline(steps=[('adding_features',\n",
       "                                              AddFeatures(where_x=0)),\n",
       "                                             ('poly', PolynomialFeatures()),\n",
       "                                             ('scaler', StandardScaler()),\n",
       "                                             ('linear_reg',\n",
       "                                              LinearRegression())]),\n",
       "             param_grid={'adding_features__functions': [[<function identity at 0x140f5a670>],\n",
       "                                                        [<function identity at 0x140f5a670>,\n",
       "                                                         <ufunc 'sin'>,\n",
       "                                                         <ufunc 'exp'>,\n",
       "                                                         <ufunc 'cos'>],\n",
       "                                                        [<ufunc 'sin'>],\n",
       "                                                        [<ufunc 'exp'>]],\n",
       "                         'poly__degree': [1, 2, 3]},\n",
       "             scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 928,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs = GridSearchCV(\n",
    "    pipeline,\n",
    "    param_grid=param_grid,\n",
    "    scoring=\"neg_mean_squared_error\",\n",
    "    cv=KFold(3, shuffle=True)\n",
    ")\n",
    "gs.fit(x2_more_points, y_more_points)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 929,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_grid = pd.DataFrame(gs.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 930,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'adding_features__functions': [<ufunc 'sin'>], 'poly__degree': 1}"
      ]
     },
     "execution_count": 930,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 931,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_grid.rename(columns = { 'mean_test_score': 'neg_mean_test_score' }, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 932,
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_adding_features__functions</th>\n",
       "      <th>param_poly__degree</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>neg_mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.002088</td>\n",
       "      <td>0.000064</td>\n",
       "      <td>0.000738</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>[&lt;ufunc 'sin'&gt;]</td>\n",
       "      <td>1</td>\n",
       "      <td>{'adding_features__functions': [&lt;ufunc 'sin'&gt;]...</td>\n",
       "      <td>-0.040405</td>\n",
       "      <td>-0.040612</td>\n",
       "      <td>-0.047605</td>\n",
       "      <td>-0.042874</td>\n",
       "      <td>0.003346</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.001898</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000752</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>[&lt;function identity at 0x140f5a670&gt;, &lt;ufunc 's...</td>\n",
       "      <td>1</td>\n",
       "      <td>{'adding_features__functions': [&lt;function iden...</td>\n",
       "      <td>-0.040724</td>\n",
       "      <td>-0.040994</td>\n",
       "      <td>-0.047577</td>\n",
       "      <td>-0.043098</td>\n",
       "      <td>0.003169</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.002207</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000740</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>[&lt;ufunc 'sin'&gt;]</td>\n",
       "      <td>2</td>\n",
       "      <td>{'adding_features__functions': [&lt;ufunc 'sin'&gt;]...</td>\n",
       "      <td>-0.040817</td>\n",
       "      <td>-0.040947</td>\n",
       "      <td>-0.047645</td>\n",
       "      <td>-0.043137</td>\n",
       "      <td>0.003189</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean_fit_time  std_fit_time  mean_score_time  std_score_time  \\\n",
       "6       0.002088      0.000064         0.000738        0.000011   \n",
       "3       0.001898      0.000028         0.000752        0.000097   \n",
       "7       0.002207      0.000123         0.000740        0.000008   \n",
       "\n",
       "                    param_adding_features__functions param_poly__degree  \\\n",
       "6                                    [<ufunc 'sin'>]                  1   \n",
       "3  [<function identity at 0x140f5a670>, <ufunc 's...                  1   \n",
       "7                                    [<ufunc 'sin'>]                  2   \n",
       "\n",
       "                                              params  split0_test_score  \\\n",
       "6  {'adding_features__functions': [<ufunc 'sin'>]...          -0.040405   \n",
       "3  {'adding_features__functions': [<function iden...          -0.040724   \n",
       "7  {'adding_features__functions': [<ufunc 'sin'>]...          -0.040817   \n",
       "\n",
       "   split1_test_score  split2_test_score  neg_mean_test_score  std_test_score  \\\n",
       "6          -0.040612          -0.047605            -0.042874        0.003346   \n",
       "3          -0.040994          -0.047577            -0.043098        0.003169   \n",
       "7          -0.040947          -0.047645            -0.043137        0.003189   \n",
       "\n",
       "   rank_test_score  \n",
       "6                1  \n",
       "3                2  \n",
       "7                3  "
      ]
     },
     "execution_count": 932,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_grid.sort_values('neg_mean_test_score', ascending=False)[:3] # 3 best performing models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 933,
   "metadata": {},
   "outputs": [],
   "source": [
    "estimator =  gs.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 934,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x140d3c3d0>]"
      ]
     },
     "execution_count": 934,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.scatter(x, y, color='blue')\n",
    "plt.plot(x, estimator.predict(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find back the sin ! :D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 935,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid_lasso = { \n",
    "    **param_grid,\n",
    "    \"linear_reg\" : [Lasso(tol=0.5)],\n",
    "    \"linear_reg__alpha\" : np.linspace(0.00000001, 1, 50)\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 936,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=3, random_state=None, shuffle=True),\n",
       "             estimator=BetterPipeline(steps=[('adding_features',\n",
       "                                              AddFeatures(where_x=0)),\n",
       "                                             ('poly', PolynomialFeatures()),\n",
       "                                             ('scaler', StandardScaler()),\n",
       "                                             ('linear_reg',\n",
       "                                              LinearRegression())]),\n",
       "             param_grid={'adding_features__functions': [[<function identity at 0x140f5a670>],\n",
       "                                                        [<function identity at 0x140f5a670>,\n",
       "                                                         <ufunc '...\n",
       "       5.71428576e-01, 5.91836739e-01, 6.12244902e-01, 6.32653065e-01,\n",
       "       6.53061228e-01, 6.73469391e-01, 6.93877554e-01, 7.14285717e-01,\n",
       "       7.34693880e-01, 7.55102043e-01, 7.75510206e-01, 7.95918369e-01,\n",
       "       8.16326532e-01, 8.36734696e-01, 8.57142859e-01, 8.77551022e-01,\n",
       "       8.97959185e-01, 9.18367348e-01, 9.38775511e-01, 9.59183674e-01,\n",
       "       9.79591837e-01, 1.00000000e+00]),\n",
       "                         'poly__degree': [1, 2, 3]},\n",
       "             scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 936,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs = GridSearchCV(pipeline, param_grid=param_grid_lasso, \n",
    "                  scoring=\"neg_mean_squared_error\", cv=KFold(3, shuffle=True))\n",
    "gs.fit(x2_more_points, y_more_points)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 938,
   "metadata": {
    "scrolled": false
   },
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_adding_features__functions</th>\n",
       "      <th>param_linear_reg</th>\n",
       "      <th>param_linear_reg__alpha</th>\n",
       "      <th>param_poly__degree</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>neg_mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>0.003711</td>\n",
       "      <td>0.000276</td>\n",
       "      <td>0.001259</td>\n",
       "      <td>0.000031</td>\n",
       "      <td>[&lt;function identity at 0x140f5a670&gt;, &lt;ufunc 's...</td>\n",
       "      <td>Lasso(alpha=1e-08, tol=0.5)</td>\n",
       "      <td>1e-08</td>\n",
       "      <td>3</td>\n",
       "      <td>{'adding_features__functions': [&lt;function iden...</td>\n",
       "      <td>-0.049015</td>\n",
       "      <td>-0.042963</td>\n",
       "      <td>-0.042469</td>\n",
       "      <td>-0.044815</td>\n",
       "      <td>0.002976</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>0.002369</td>\n",
       "      <td>0.000276</td>\n",
       "      <td>0.000835</td>\n",
       "      <td>0.000136</td>\n",
       "      <td>[&lt;function identity at 0x140f5a670&gt;, &lt;ufunc 's...</td>\n",
       "      <td>Lasso(alpha=1e-08, tol=0.5)</td>\n",
       "      <td>1e-08</td>\n",
       "      <td>2</td>\n",
       "      <td>{'adding_features__functions': [&lt;function iden...</td>\n",
       "      <td>-0.053484</td>\n",
       "      <td>-0.045553</td>\n",
       "      <td>-0.044973</td>\n",
       "      <td>-0.048004</td>\n",
       "      <td>0.003883</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>0.003686</td>\n",
       "      <td>0.000721</td>\n",
       "      <td>0.001281</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>[&lt;function identity at 0x140f5a670&gt;, &lt;ufunc 's...</td>\n",
       "      <td>Lasso(alpha=1e-08, tol=0.5)</td>\n",
       "      <td>0.0204082</td>\n",
       "      <td>3</td>\n",
       "      <td>{'adding_features__functions': [&lt;function iden...</td>\n",
       "      <td>-0.055127</td>\n",
       "      <td>-0.046499</td>\n",
       "      <td>-0.045358</td>\n",
       "      <td>-0.048995</td>\n",
       "      <td>0.004361</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     mean_fit_time  std_fit_time  mean_score_time  std_score_time  \\\n",
       "152       0.003711      0.000276         0.001259        0.000031   \n",
       "151       0.002369      0.000276         0.000835        0.000136   \n",
       "155       0.003686      0.000721         0.001281        0.000088   \n",
       "\n",
       "                      param_adding_features__functions  \\\n",
       "152  [<function identity at 0x140f5a670>, <ufunc 's...   \n",
       "151  [<function identity at 0x140f5a670>, <ufunc 's...   \n",
       "155  [<function identity at 0x140f5a670>, <ufunc 's...   \n",
       "\n",
       "                param_linear_reg param_linear_reg__alpha param_poly__degree  \\\n",
       "152  Lasso(alpha=1e-08, tol=0.5)                   1e-08                  3   \n",
       "151  Lasso(alpha=1e-08, tol=0.5)                   1e-08                  2   \n",
       "155  Lasso(alpha=1e-08, tol=0.5)               0.0204082                  3   \n",
       "\n",
       "                                                params  split0_test_score  \\\n",
       "152  {'adding_features__functions': [<function iden...          -0.049015   \n",
       "151  {'adding_features__functions': [<function iden...          -0.053484   \n",
       "155  {'adding_features__functions': [<function iden...          -0.055127   \n",
       "\n",
       "     split1_test_score  split2_test_score  neg_mean_test_score  \\\n",
       "152          -0.042963          -0.042469            -0.044815   \n",
       "151          -0.045553          -0.044973            -0.048004   \n",
       "155          -0.046499          -0.045358            -0.048995   \n",
       "\n",
       "     std_test_score  rank_test_score  \n",
       "152        0.002976                1  \n",
       "151        0.003883                2  \n",
       "155        0.004361                3  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "BetterPipeline(steps=[('adding_features',\n",
       "                       AddFeatures(functions=[<function identity at 0x140f5a670>,\n",
       "                                              <ufunc 'sin'>, <ufunc 'exp'>,\n",
       "                                              <ufunc 'cos'>],\n",
       "                                   where_x=0)),\n",
       "                      ('poly', PolynomialFeatures(degree=3)),\n",
       "                      ('scaler', StandardScaler()),\n",
       "                      ('linear_reg', Lasso(alpha=1e-08, tol=0.5))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x14102a2b0>]"
      ]
     },
     "execution_count": 938,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "df_grid = pd.DataFrame(gs.cv_results_)\n",
    "df_grid.rename(columns = { 'mean_test_score': 'neg_mean_test_score' }, inplace=True)\n",
    "display(df_grid.sort_values('neg_mean_test_score', ascending=False)[:3]) # 3 best performing models\n",
    "estimator =  gs.best_estimator_\n",
    "display(estimator)\n",
    "plt.scatter(x, y, color='blue')\n",
    "plt.plot(x, estimator.predict(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lasso might have canceled out some params i guess (at least on the 3rd model)"
   ]
  }
 ],
 "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": "3.8.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}