ogrisel/minimum_norm_ridge_limit.ipynb

## minimum_norm_ridge_limit.ipynb
{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ridge objective parametrizations and intercept handling\n",
    "\n",
    "The purpose of this notebook is two empirically study two ways to parametrize the Ridge regression problem, with intercept and study the impact with low (and high!) regularization on underdetermined problems.\n",
    "\n",
    "- **Type \"a\" ridge** is the one currently implemented in scikit-learn: it fits a **ridge estimator (without intercept) on centered X and y** and computes the intercept from the optimal coef a-posteriori (analytically): `intercept = y_train.mean() - X_train.mean(axis=0) @ coef`. \n",
    "\n",
    "- **Type \"b\" ridge horizontally concatenates a constant column of 1 to X**, and the solve for the ridge using standard least squares with Tikhonov diagonal damping except that the last diagonal element is set to zero to avoid penalizing the intercept.\n",
    "\n",
    "Both formulations fit a ridge estimator with unpenalized intercept on the original problem and as this notebook shows, both converge to the same solution when alpha is large enough. However, on underdetermined problem their behavior diverge slightly and converge to different minimum norm solution of the OLS problem in the `alpha -> 0` limit:\n",
    "\n",
    "- **Type \"a\" OLS does not include the intercept** parameter in the computation of the norm used for the minimum norm objective.\n",
    "- **Type \"b\" OLS includes the intercept** parameter in the computation of the norm used for the minimum norm objective.\n",
    "\n",
    "This notebook serves as a guiding experiment to make a decision for the design of scikit-learn:\n",
    "\n",
    "- https://github.com/scikit-learn/scikit-learn/pull/25948\n",
    "\n",
    "As we can see in the following, the discrepency between the 2 types of ridge appears for non-zero values of alpha, as soon as one of the 2 underlying linear algebra system is no longer numerically determined."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200.66049258604295, 522.6463231353866)"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.datasets import make_regression\n",
    "from numpy.testing import assert_allclose\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "rng = np.random.RandomState(0)\n",
    "\n",
    "# User a large data set to be able many CV iterations from the same\n",
    "# distribution and average the results.\n",
    "n_samples_all, n_features = 10_000, 100\n",
    "n_samples_train = 50\n",
    "\n",
    "# Use a large intercept in the data generative process because the two\n",
    "# approaches are different only when fit_intercept=True and matters.\n",
    "intercept_true = 200.0\n",
    "\n",
    "X, y_noise_free, coef_true = make_regression(\n",
    "    n_samples=n_samples_all,\n",
    "    n_features=n_features,\n",
    "    n_informative=int(0.8 * n_features),\n",
    "    noise=0.0,\n",
    "    coef=True,\n",
    "    bias=intercept_true,\n",
    "    random_state=rng,\n",
    ")\n",
    "# Quick sanity check.\n",
    "assert_allclose(X @ coef_true + intercept_true, y_noise_free)\n",
    "\n",
    "# # Add some noise manually to the target to make the generalization problem less easy.\n",
    "y = y_noise_free + y_noise_free.std() / 10 * rng.normal(size=n_samples_all)\n",
    "\n",
    "y.mean(), y.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.hist(y, bins=30);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the intercept is large enough compared to the range of variation of y. This study requires it because we want to study the impact of intercept estimation by both rige types.\n",
    "\n",
    "Let's now implement both types of ridge estimators the traditional scikit-learn API:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from numpy.linalg import pinv\n",
    "from scipy.linalg import lstsq\n",
    "\n",
    "\n",
    "class MinimalRidge(LinearRegression):\n",
    "    def __init__(\n",
    "        self, alpha=1.0, ridge_type=\"a\", solver=\"pinv\", fit_intercept=True, rcond=1e-12\n",
    "    ):\n",
    "        self.alpha = alpha\n",
    "        self.fit_intercept = fit_intercept\n",
    "        self.rcond = rcond\n",
    "        self.ridge_type = ridge_type\n",
    "        self.solver = solver\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        assert self.fit_intercept, \"This experiment only deals with fit_intercept=True\"\n",
    "        if self.ridge_type == \"a\":\n",
    "            return self._fit_a(X, y)\n",
    "        elif self.ridge_type == \"b\":\n",
    "            return self._fit_b(X, y)\n",
    "        else:\n",
    "            raise ValueError(f\"Unknown ridge_type: {self.ridge_type}\")\n",
    "\n",
    "    def _fit_a(self, X, y):\n",
    "        # Center X and y and compute the intercept the intercept from the\n",
    "        # solution of the ridge on the centered data. When alpha=0, this\n",
    "        # converges to the minimum norm OLS where the intercept is not\n",
    "        # included in the computation of the norm of the solution.\n",
    "        X_m = X.mean(axis=0)\n",
    "        X_c = X - X_m\n",
    "        y_m = y.mean()\n",
    "        y_c = y - y_m\n",
    "        damping_matrix = self.alpha * np.eye(X_c.shape[1])\n",
    "        if self.solver == \"pinv\":\n",
    "            self.coef_ = (\n",
    "                pinv(X_c.T @ X_c + damping_matrix, rcond=self.rcond) @ X_c.T @ y_c\n",
    "            )\n",
    "        elif self.solver == \"lstsq\":\n",
    "            self.coef_ = lstsq(\n",
    "                np.vstack([X_c, damping_matrix]),\n",
    "                np.concatenate([y_c, np.zeros(X.shape[1])]),\n",
    "                cond=self.rcond,\n",
    "            )[0]\n",
    "        else:\n",
    "            raise ValueError(f\"Unknown solver: {self.solver}\")\n",
    "        # Compute the intercept from the solution of the ridge on the centered\n",
    "        self.intercept_ = y_m - X_m @ self.coef_\n",
    "        return self\n",
    "\n",
    "    def _fit_b(self, X, y):\n",
    "        # Solve for the ridge problem by concatenating the data with a column\n",
    "        # of 1. Use a damping matrix with a zero value for the last diagonal\n",
    "        # element (unpenalized intercept). Note when alpha is zero, this\n",
    "        # converges to the minimum norm OLS problem where the intercept is\n",
    "        # included in computing the norm of the solution.\n",
    "\n",
    "        n_samples, n_features = X.shape\n",
    "        X_with_ones = np.hstack([X, np.ones((n_samples, 1))])\n",
    "        damping_matrix = self.alpha * np.eye(n_features + 1)\n",
    "        damping_matrix[-1, -1] = 0  # unpenalized intercept\n",
    "        if self.solver == \"pinv\":\n",
    "            w = (\n",
    "                pinv(X_with_ones.T @ X_with_ones + damping_matrix, rcond=self.rcond)\n",
    "                @ X_with_ones.T\n",
    "                @ y\n",
    "            )\n",
    "        elif self.solver == \"lstsq\":\n",
    "            w = lstsq(\n",
    "                np.vstack([X_with_ones, damping_matrix[:-1]]),\n",
    "                np.concatenate([y, np.zeros(n_features)]),\n",
    "                cond=self.rcond,\n",
    "            )[0]\n",
    "        else:\n",
    "            raise ValueError(f\"Unknown solver: {self.solver}\")\n",
    "        self.coef_ = w[:-1]\n",
    "        self.intercept_ = w[-1]\n",
    "        return self\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fixed split to run some sanity checks on a small training set.\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, random_state=0, train_size=n_samples_train\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sanity check #0, check that both models can perfectly fit the training data\n",
    "# with no regularization as n_samples_train is small.\n",
    "assert_allclose(MinimalRidge(alpha=0.0, ridge_type=\"a\").fit(X_train, y_train).score(X_train, y_train), 1.0)\n",
    "assert_allclose(MinimalRidge(alpha=0.0, ridge_type=\"b\").fit(X_train, y_train).score(X_train, y_train), 1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sanity check #1, check that both models can reach the expected high\n",
    "# penalization limit model, and in particular that the intercept is effectively\n",
    "# not penalized.\n",
    "for solver in [\"pinv\", \"lstsq\"]:\n",
    "    ridge_a = MinimalRidge(alpha=1e12, ridge_type=\"a\", solver=solver).fit(X_train, y_train)\n",
    "    assert_allclose(ridge_a.coef_, 0.0, atol=1e-7)\n",
    "    assert_allclose(ridge_a.intercept_, y_train.mean())\n",
    "\n",
    "    ridge_a = MinimalRidge(alpha=1e30, ridge_type=\"a\", solver=solver).fit(X_train, y_train)\n",
    "    assert_allclose(ridge_a.coef_, 0.0, atol=1e-12)\n",
    "    assert_allclose(ridge_a.intercept_, y_train.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "for solver in [\"pinv\", \"lstsq\"]:\n",
    "    ridge_b = MinimalRidge(alpha=1e12, ridge_type=\"b\", solver=solver)\n",
    "    ridge_b.fit(X_train, y_train)\n",
    "    assert_allclose(ridge_b.coef_, 0.0, atol=1e-7)\n",
    "    assert_allclose(ridge_b.intercept_, y_train.mean(), rtol=1e-6)\n",
    "\n",
    "    ridge_b = MinimalRidge(alpha=1e30, ridge_type=\"b\", solver=solver)\n",
    "    ridge_b.fit(X_train, y_train)\n",
    "    assert_allclose(ridge_b.coef_, 0.0, atol=1e-12)\n",
    "\n",
    "    # XXX: the following unexpectedly fails: the intercept stays very close to 0\n",
    "    # instead of being the mean of the target. Making it work would probably\n",
    "    # require to finely adjust rcond to the value of alpha...\n",
    "\n",
    "    # assert_allclose(ridge_b.intercept_, y_train.mean())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "solver = 'pinv' - alpha = 0.0e+00: coef_diff = 4.0e-09 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e-03: coef_diff = 1.8e-08 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e-02: coef_diff = 2.1e-09 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e-01: coef_diff = 2.2e-10 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e+00: coef_diff = 3.3e-10 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e+01: coef_diff = 3.8e-10 - PASS\n",
      "solver = 'pinv' - alpha = 1.0e+02: coef_diff = 4.3e-10 - PASS\n",
      "solver = 'lstsq' - alpha = 0.0e+00: coef_diff = 4.0e-09 - PASS\n",
      "solver = 'lstsq' - alpha = 1.0e-03: coef_diff = 1.1e-02 - FAIL\n",
      "solver = 'lstsq' - alpha = 1.0e-02: coef_diff = 1.1e-01 - FAIL\n",
      "solver = 'lstsq' - alpha = 1.0e-01: coef_diff = 9.7e-01 - FAIL\n",
      "solver = 'lstsq' - alpha = 1.0e+00: coef_diff = 3.3e-10 - PASS\n",
      "solver = 'lstsq' - alpha = 1.0e+01: coef_diff = 1.5e+02 - FAIL\n",
      "solver = 'lstsq' - alpha = 1.0e+02: coef_diff = 1.4e+02 - FAIL\n"
     ]
    }
   ],
   "source": [
    "# Sanity check #2, check that Ridge type \"a\" matches scikit-learn's Ridge\n",
    "from sklearn.linear_model import Ridge\n",
    "\n",
    "for solver in [\"pinv\", \"lstsq\"]:\n",
    "    for alpha in [0., 0.001, 0.01, 0.1, 1., 10., 100.]:\n",
    "        coef_diff = np.linalg.norm(\n",
    "            MinimalRidge(alpha=alpha, ridge_type=\"a\", solver=solver).fit(X_train, y_train).coef_ -\n",
    "            Ridge(alpha=alpha, solver=\"lsqr\", tol=1e-12).fit(X_train, y_train).coef_\n",
    "        )\n",
    "        print(f\"{solver = } - {alpha = :0.1e}: {coef_diff = :0.1e}\", end=\"\")\n",
    "        if coef_diff < 1e-5:\n",
    "            print(\" - PASS\")\n",
    "        else:\n",
    "            print(\" - FAIL\")\n",
    "\n",
    "# XXX: The lstsq solver is not as stable as the pinv solver. Or maybe the lsqr\n",
    "# way we use the lsqr solver in scikit-learn is the culprit?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Split #0\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a8592eb26feb4f0084cd826d4aec7fbd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "376733fab643474b8f96168fc00de0f7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Split #1\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e1c204700b974e4dbb276bf33ea342a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "470dd6efd8e04138b2350f23d34667cc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Split #2\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f6453a65f62b47fea0a3d3aa84bb2591",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "87b97068fcb844dabfb9285f733ce3c9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/300 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "# The rest of the analysis will be conducted for a single solver. Feel free to\n",
    "# comment/uncomment the definition of the `solver` parameter, restart and run\n",
    "# all cells to check that the conclusions are robust to the choice of the\n",
    "# solver.\n",
    "\n",
    "#  solver = \"lstsq\"\n",
    "solver = \"pinv\"\n",
    "\n",
    "info = []\n",
    "for split_idx in range(3):\n",
    "    print(f\"Split #{split_idx}\")\n",
    "    X_train, X_test, y_train, y_test = train_test_split(\n",
    "        X, y, random_state=split_idx, train_size=n_samples_train\n",
    "    )\n",
    "\n",
    "    for ridge_type in [\"a\", \"b\"]:\n",
    "        ols = MinimalRidge(alpha=0, ridge_type=ridge_type, solver=solver)\n",
    "        ols.fit(X_train, y_train)\n",
    "        ols_params = np.concatenate([ols.coef_, [ols.intercept_]])\n",
    "        ols_test_preds = ols.predict(X_test)\n",
    "\n",
    "        for alpha in tqdm(np.logspace(-20, 20, 300)):\n",
    "            ridge = MinimalRidge(alpha=alpha, ridge_type=ridge_type, solver=solver)\n",
    "            ridge.fit(X_train, y_train)\n",
    "            ridge_params = np.concatenate([ridge.coef_, [ridge.intercept_]])\n",
    "            info.append(\n",
    "                {\n",
    "                    \"split_idx\": split_idx,\n",
    "                    \"ridge_type\": ridge_type,\n",
    "                    \"alpha\": alpha,\n",
    "                    \"coef\": ridge.coef_,\n",
    "                    \"intercept\": ridge.intercept_,\n",
    "                    \"params\": ridge_params,\n",
    "                    \"param_dist_to_ols\": np.linalg.norm(ridge_params - ols_params),\n",
    "                    \"pred_dist_to_ols\": np.linalg.norm(ridge.predict(X_test) - ols_test_preds),\n",
    "                    \"train_mse\": mean_squared_error(y_train, ridge.predict(X_train)),\n",
    "                    \"train_r2\": ridge.score(X_train, y_train),\n",
    "                    \"test_mse\": mean_squared_error(y_test, ridge.predict(X_test)),\n",
    "                    \"test_r2\": ridge.score(X_test, y_test),\n",
    "                }\n",
    "            )\n",
    "    print()\n",
    "info = pd.DataFrame(info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "metrics_plot_params = [\n",
    "    # {\"metric\": \"train_mse\", \"ylabel\": \"Train MSE\", \"yscale\": \"log\"},\n",
    "    # {\"metric\": \"test_mse\", \"ylabel\": \"Test MSE\", \"yscale\": \"log\"},\n",
    "    {\"metric\": \"train_r2\", \"ylabel\": \"Train R2\", \"yscale\": \"linear\"},\n",
    "    {\"metric\": \"test_r2\", \"ylabel\": \"Test R2\", \"yscale\": \"linear\"},\n",
    "    {\"metric\": \"param_dist_to_ols\", \"ylabel\": \"Param dist to OLS\", \"yscale\": \"linear\"},\n",
    "    {\"metric\": \"pred_dist_to_ols\", \"ylabel\": \"Pred dist to OLS\", \"yscale\": \"linear\"},\n",
    "    {\"metric\": \"intercept\", \"ylabel\": \"Intercept\", \"yscale\": \"linear\"},\n",
    "]\n",
    "\n",
    "for split_idx, split_info in info.groupby(\"split_idx\"):\n",
    "    fig, axes = plt.subplots(nrows=len(metrics_plot_params), figsize=(12, 8), sharex=True)\n",
    "    axes[0].set_xscale('log')  # alpha in log space\n",
    "    fig.suptitle(f\"Split #{split_idx}\")\n",
    "    for metric_plot_params, ax in zip(metrics_plot_params, axes):\n",
    "        for ridge_type in [\"a\", \"b\"]:\n",
    "            ax.set(ylabel=metric_plot_params[\"ylabel\"], yscale=metric_plot_params.get(\"yscale\", \"linear\"))\n",
    "            split_info.query(\"ridge_type == @ridge_type\").plot(x=\"alpha\", y=metric_plot_params[\"metric\"], label=ridge_type, ax=ax)\n",
    "\n",
    "plt.legend();"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analysis\n",
    "\n",
    "- Both type a and type b ridge formulations converge to the same solution for a wide range of intermediate values of alpha (explicitly checked in the end of this notebook).\n",
    "- Both type a and type b ridge formulations reach their own OLS limit (type a or type b), however the **type \"b\" formulation fails to continuously progress to the type \"b\" minimal norm OLS solution**: indeed we can always observe a discontinuity when alpha is around 1e-9...\n",
    "- Type a and type b OLS are different (as expected) and in general, type a OLS seems to be more correct (lower test MSE / higher test R2) more often (but it depends on the sampling of the training set).\n",
    "- More suprisingly, type b ridge also has a numerical discontinuity towards infinitely regularized models: it fails to converge to the constant `y_train.mean()` prediction because of numerical unstability in the estimation of the intercept: for highly regularization models, this estimator finds near-zero intercept values instead... This is completely wrong.\n",
    "- Both discontinuities of the type \"b\" ridge formulation seem to stem from numerically unstable estimation of the intercept at low or high alpha values as can be seen on the last row of the above plot.\n",
    "\n",
    "- [x] TODO: confirm that those conclusions still hold with alternative numerical solvers (e.g. `scipy.linalg.lstsq` with properly set `cond` parameter to threshold the near zero singular values appropriately as we do with the current `pinv` based solver that also relies on an SVD solver). => done: the previous conclusions are the same, although the `lstsq`-based implemetnation does not seem as stable as the `pinv` based implementation (as reported in the sanity checks).\n",
    "\n",
    "- [ ] TODO: but even more work: implement both ridge types with `scipy.sparse.linalg.lsqr` or `scipy.sparse.linalg.lsmr` instead.\n",
    "\n",
    "\n",
    "In the following we check that both ridge estimator converge to the same solution for intermediate values of alpha:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "info_a = (\n",
    "    info.query(\"ridge_type == 'a'\").sort_values([\"split_idx\", \"alpha\"]).reset_index()\n",
    ")\n",
    "info_b = (\n",
    "    info.query(\"ridge_type == 'b'\").sort_values([\"split_idx\", \"alpha\"]).reset_index()\n",
    ")\n",
    "\n",
    "param_diff = np.vstack(info_a[\"params\"] - info_b[\"params\"])\n",
    "info_a[\"param_diff_a_to_b\"] = (param_diff**2).sum(axis=1)\n",
    "\n",
    "for split_idx, split_info in info_a.groupby(\"split_idx\"):\n",
    "    split_info.plot(\n",
    "        x=\"alpha\",\n",
    "        y=\"param_diff_a_to_b\",\n",
    "        logx=True,\n",
    "        figsize=(12, 4),\n",
    "        title=f\"Split #{split_idx}\",\n",
    "        ylabel=\"Param diff (a to b)\",\n",
    "    )\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dev",
   "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": "3.11.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}