Finish creating and saving all the plots
This commit is contained in:
+92
-197
@@ -18,15 +18,14 @@
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import optimizers\n",
|
||||
"import datamanip\n",
|
||||
"import plotting\n",
|
||||
"import pyoptim.optimizers as optimizers\n",
|
||||
"import pyoptim.datamanip as datamanip\n",
|
||||
"import pyoptim.plotting as plotting\n",
|
||||
"\n",
|
||||
"import numpy as np\n",
|
||||
"import matplotlib.pyplot as plt\n",
|
||||
"from sklearn.model_selection import train_test_split\n",
|
||||
"import os\n",
|
||||
"import time"
|
||||
"import os"
|
||||
]
|
||||
},
|
||||
{
|
||||
@@ -37,7 +36,7 @@
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"FIG_DIR = os.path.abspath(\n",
|
||||
" os.path.join(os.path.dirname(plotting.__file__), \"..\", \"figures\")\n",
|
||||
" os.path.join(os.path.dirname(plotting.__file__), \"../..\", \"figures\")\n",
|
||||
")\n",
|
||||
"print(FIG_DIR)"
|
||||
]
|
||||
@@ -53,12 +52,7 @@
|
||||
"y = datamanip.noise_data(datamanip.runge_function(x), 1.0)\n",
|
||||
"x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2)\n",
|
||||
"\n",
|
||||
"fig, ax = plt.subplots(figsize=plotting.get_figsize(0.5))\n",
|
||||
"ax.scatter(x_train, y_train, label=\"Train Data\", s=1)\n",
|
||||
"ax.scatter(x_test, y_test, label=\"Test Data\", s=1)\n",
|
||||
"ax.set_xlabel(\"$x$\")\n",
|
||||
"ax.set_ylabel(\"$y$\")\n",
|
||||
"ax.legend()\n",
|
||||
"fig, ax = plotting.scatter_dataset(x_train, x_test, y_train, y_test)\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"data_scatter.png\"), dpi=300)"
|
||||
]
|
||||
},
|
||||
@@ -273,29 +267,30 @@
|
||||
"fig, (ax1, ax2) = plt.subplots(1, 2, figsize=plotting.get_figsize(0.5))\n",
|
||||
"num_iters = 1_000\n",
|
||||
"\n",
|
||||
"for learning_rate in np.logspace(-4, 0, 5):\n",
|
||||
" OLSGrad_Desc = optimizers.OLSGradientDescent(\n",
|
||||
" learning_rate=learning_rate, num_iterations=num_iters\n",
|
||||
"learning_rates = np.logspace(-4, 0, 5)\n",
|
||||
"ols_optimizers = [\n",
|
||||
" (\n",
|
||||
" optimizers.OLSGradientDescent,\n",
|
||||
" {\"learning_rate\": learning_rate, \"num_iterations\": num_iters},\n",
|
||||
" )\n",
|
||||
" OLSGrad_Desc.fit(X_tr, y_tr)\n",
|
||||
" cost_history = OLSGrad_Desc.cost_history\n",
|
||||
" RidgeGrad_Desc = optimizers.RidgeGradientDescent(\n",
|
||||
" learning_rate=learning_rate, num_iterations=num_iters, lam=0.1\n",
|
||||
" for learning_rate in learning_rates\n",
|
||||
"]\n",
|
||||
"ridge_optimizers = [\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeGradientDescent,\n",
|
||||
" {\"learning_rate\": learning_rate, \"num_iterations\": num_iters, \"lam\": 0.1},\n",
|
||||
" )\n",
|
||||
" RidgeGrad_Desc.fit(X_tr, y_tr)\n",
|
||||
" cost_history_ridge = RidgeGrad_Desc.cost_history\n",
|
||||
" ax1.plot(cost_history, label=f\"$\\\\eta = 10^{{{int(np.log10(learning_rate))}}}$\")\n",
|
||||
" ax2.plot(\n",
|
||||
" cost_history_ridge, label=f\"$\\\\eta = 10^{{{int(np.log10(learning_rate))}}}$\"\n",
|
||||
" )\n",
|
||||
"ax1.set_xlabel(\"Iteration\")\n",
|
||||
"ax1.set_ylabel(\"OLS Cost\")\n",
|
||||
"ax1.legend()\n",
|
||||
" for learning_rate in learning_rates\n",
|
||||
"]\n",
|
||||
"labels = [\n",
|
||||
" f\"$\\\\eta = 10^{{{int(np.log10(learning_rate))}}}$\"\n",
|
||||
" for learning_rate in learning_rates\n",
|
||||
"]\n",
|
||||
"\n",
|
||||
"plotting.plot_optimizers(ax1, ols_optimizers, X_tr, y_tr, \"OLS Cost\", labels=labels)\n",
|
||||
"plotting.plot_optimizers(ax2, ridge_optimizers, X_tr, y_tr, \"Ridge Cost\", labels=labels)\n",
|
||||
"ax1.set_ylim(bottom=0.465, top=0.505)\n",
|
||||
"ax2.set_ylim(bottom=0.4775, top=0.505)\n",
|
||||
"ax2.set_xlabel(\"Iteration\")\n",
|
||||
"ax2.set_ylabel(\"Ridge Cost\")\n",
|
||||
"ax2.legend()\n",
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"gradient_descent_convergence.pdf\"))"
|
||||
]
|
||||
@@ -306,20 +301,6 @@
|
||||
"id": "13",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"def get_cost_history(optimizer, X, y, **kwargs):\n",
|
||||
" opt = optimizer(**kwargs)\n",
|
||||
" opt.fit(X, y)\n",
|
||||
" iterations = opt.get_epochs()\n",
|
||||
" return iterations, opt.cost_history"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "14",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"fig, (ax1, ax2) = plt.subplots(2, 1, figsize=plotting.get_figsize(0.8))\n",
|
||||
"\n",
|
||||
@@ -333,87 +314,36 @@
|
||||
"learning_rate_ridge = 0.01\n",
|
||||
"lam = 0.1\n",
|
||||
"\n",
|
||||
"general_kwargs = {\n",
|
||||
" \"num_iterations\": num_iters,\n",
|
||||
"}\n",
|
||||
"optimizer_kwargs = [\n",
|
||||
" {},\n",
|
||||
" {\"delta\": 0.9},\n",
|
||||
" {},\n",
|
||||
" {\"gamma\": 0.995},\n",
|
||||
" {\"beta1\": 0.9, \"beta2\": 0.999},\n",
|
||||
"]\n",
|
||||
"ols_kwargs = {\"learning_rate\": learning_rate_ols, **general_kwargs}\n",
|
||||
"ridge_kwargs = {\n",
|
||||
" \"learning_rate\": learning_rate_ridge,\n",
|
||||
" \"lam\": lam,\n",
|
||||
" **general_kwargs,\n",
|
||||
"}\n",
|
||||
"optimizers_ols = [\n",
|
||||
" (\n",
|
||||
" optimizers.OLSGradientDescent,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ols},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.OLSMomentum,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ols, \"delta\": 0.9},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.OLSAdaGrad,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ols},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.OLSRMSProp,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ols, \"gamma\": 0.9},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.OLSAdam,\n",
|
||||
" {\n",
|
||||
" \"num_iterations\": num_iters,\n",
|
||||
" \"learning_rate\": learning_rate_ols,\n",
|
||||
" \"beta1\": 0.9,\n",
|
||||
" \"beta2\": 0.999,\n",
|
||||
" },\n",
|
||||
" ),\n",
|
||||
" (opt, {**ols_kwargs, **opt_kwargs})\n",
|
||||
" for opt, opt_kwargs in zip(optimizers.OLS_GD_OPTIMIZERS, optimizer_kwargs)\n",
|
||||
"]\n",
|
||||
"\n",
|
||||
"optimizers_ridge = [\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeGradientDescent,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ridge, \"lam\": lam},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeMomentum,\n",
|
||||
" {\n",
|
||||
" \"num_iterations\": num_iters,\n",
|
||||
" \"learning_rate\": learning_rate_ridge,\n",
|
||||
" \"delta\": 0.9,\n",
|
||||
" \"lam\": lam,\n",
|
||||
" },\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeAdaGrad,\n",
|
||||
" {\"num_iterations\": num_iters, \"learning_rate\": learning_rate_ridge, \"lam\": lam},\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeRMSProp,\n",
|
||||
" {\n",
|
||||
" \"num_iterations\": num_iters,\n",
|
||||
" \"learning_rate\": learning_rate_ridge,\n",
|
||||
" \"gamma\": 0.9,\n",
|
||||
" \"lam\": lam,\n",
|
||||
" },\n",
|
||||
" ),\n",
|
||||
" (\n",
|
||||
" optimizers.RidgeAdam,\n",
|
||||
" {\n",
|
||||
" \"num_iterations\": num_iters,\n",
|
||||
" \"learning_rate\": learning_rate_ridge,\n",
|
||||
" \"beta1\": 0.9,\n",
|
||||
" \"beta2\": 0.999,\n",
|
||||
" \"lam\": lam,\n",
|
||||
" },\n",
|
||||
" ),\n",
|
||||
" (opt, {**ridge_kwargs, **opt_kwargs})\n",
|
||||
" for opt, opt_kwargs in zip(optimizers.RIDGE_GD_OPTIMIZERS, optimizer_kwargs)\n",
|
||||
"]\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"def plot_optimizers(ax, optimizer_list, X, y, ylabel=\"Cost\"):\n",
|
||||
" for opt_class, params in optimizer_list:\n",
|
||||
" iter, history = get_cost_history(opt_class, X, y, **params)\n",
|
||||
" ax.plot(iter, history, label=f\"{opt_class.__name__}\")\n",
|
||||
" ax.set_xlabel(\"Iteration\")\n",
|
||||
" ax.set_ylabel(ylabel)\n",
|
||||
" ax.legend()\n",
|
||||
"plotting.plot_optimizers(ax1, optimizers_ols, X_tr, y_tr, ylabel=\"Cost (OLS)\")\n",
|
||||
"plotting.plot_optimizers(ax2, optimizers_ridge, X_tr, y_tr, ylabel=\"Cost (Ridge)\")\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"plot_optimizers(ax1, optimizers_ols, X_tr, y_tr, ylabel=\"Cost (OLS)\")\n",
|
||||
"ax1.set_ylim(bottom=0.465, top=0.505)\n",
|
||||
"\n",
|
||||
"plot_optimizers(ax2, optimizers_ridge, X_tr, y_tr, ylabel=\"Cost (Ridge)\")\n",
|
||||
"ax2.set_ylim(bottom=0.4775, top=0.505)\n",
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"optimizer_comparison.pdf\"))"
|
||||
@@ -422,7 +352,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "15",
|
||||
"id": "14",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -464,7 +394,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "16",
|
||||
"id": "15",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -478,6 +408,7 @@
|
||||
" learning_rate=0.1,\n",
|
||||
" num_iterations=num_epochs * batches_per_epoch,\n",
|
||||
" batch_size=batch_size,\n",
|
||||
" batches_per_epoch=batches_per_epoch,\n",
|
||||
")\n",
|
||||
"ridge_gd = optimizers.RidgeGradientDescent(\n",
|
||||
" learning_rate=0.01, lam=0.1, num_iterations=num_epochs\n",
|
||||
@@ -487,6 +418,7 @@
|
||||
" lam=0.1,\n",
|
||||
" num_iterations=num_epochs * batches_per_epoch,\n",
|
||||
" batch_size=batch_size,\n",
|
||||
" batches_per_epoch=batches_per_epoch,\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"x = np.linspace(-1, 1, X_size)\n",
|
||||
@@ -498,50 +430,28 @@
|
||||
"X_tr, X_te = datamanip.scale_data(X_train, X_test)\n",
|
||||
"y_tr, y_te = datamanip.scale_data(y_train, y_test)\n",
|
||||
"\n",
|
||||
"fig, ax = plt.subplots(figsize=plotting.get_figsize(0.5))\n",
|
||||
"\n",
|
||||
"for optimizer, label in zip(\n",
|
||||
" [ols_sgd, ols_gd, ridge_sgd, ridge_gd],\n",
|
||||
" [\n",
|
||||
" \"Stochastic GD\",\n",
|
||||
" \"Gradient Descent\",\n",
|
||||
" \"Ridge Stochastic GD\",\n",
|
||||
" \"Ridge Gradient Descent\",\n",
|
||||
" ],\n",
|
||||
"):\n",
|
||||
" start_time = time.time()\n",
|
||||
" optimizer.fit(X_tr, y_tr)\n",
|
||||
" time_taken = time.time() - start_time\n",
|
||||
" cost_history = optimizer._cost_history\n",
|
||||
" if \"Stochastic\" in label:\n",
|
||||
" epochs = np.arange(1, num_epochs * batches_per_epoch + 1) / batches_per_epoch\n",
|
||||
" ls = \"-\"\n",
|
||||
" else:\n",
|
||||
" epochs = np.arange(1, num_epochs + 1)\n",
|
||||
" ls = \"--\"\n",
|
||||
" c = \"C0\" if \"Ridge\" not in label else \"C1\"\n",
|
||||
" ax.plot(\n",
|
||||
" epochs,\n",
|
||||
" cost_history,\n",
|
||||
" label=f\"{label} ({time_taken:.2f} s)\",\n",
|
||||
" linestyle=ls,\n",
|
||||
" color=c,\n",
|
||||
" )\n",
|
||||
"ax.set_xlabel(\"Epoch\")\n",
|
||||
"ax.set_ylabel(\"Cost\")\n",
|
||||
"ax.legend()\n",
|
||||
"# ax.set_ylim(bottom=0.0, top=0.505)\n",
|
||||
"fig.tight_layout()"
|
||||
"optimizer_list = [ols_sgd, ols_gd, ridge_sgd, ridge_gd]\n",
|
||||
"labels = [\n",
|
||||
" \"Stochastic GD\",\n",
|
||||
" \"Gradient Descent\",\n",
|
||||
" \"Ridge Stochastic GD\",\n",
|
||||
" \"Ridge Gradient Descent\",\n",
|
||||
"]\n",
|
||||
"fig, ax = plotting.optimization_performance_evaluation(\n",
|
||||
" optimizer_list, labels, X_tr, y_tr\n",
|
||||
")\n",
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"optimization_performance.pdf\"))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "17",
|
||||
"id": "16",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"X_size = 10_000\n",
|
||||
"X_size = 1_000_000\n",
|
||||
"num_epochs = 1000\n",
|
||||
"batches_per_epoch = 100\n",
|
||||
"batch_size = 128\n",
|
||||
@@ -587,67 +497,52 @@
|
||||
"optimizers_ols = [\n",
|
||||
" (opt, {**kwargs, **ols_kwargs})\n",
|
||||
" for opt, kwargs in zip(\n",
|
||||
" [\n",
|
||||
" optimizers.OLSStochasticGradientDescent,\n",
|
||||
" optimizers.OLSMomentumSGD,\n",
|
||||
" optimizers.OLSAdaGradSGD,\n",
|
||||
" optimizers.OLSRMSPropSGD,\n",
|
||||
" optimizers.OLSAdamSGD,\n",
|
||||
" ],\n",
|
||||
" optimizers.OLS_SGD_OPTIMIZERS,\n",
|
||||
" optimizer_kwargs,\n",
|
||||
" )\n",
|
||||
"]\n",
|
||||
"optimizers_ridge = [\n",
|
||||
" (opt, {**kwargs, **ridge_kwargs})\n",
|
||||
" for opt, kwargs in zip(\n",
|
||||
" [\n",
|
||||
" optimizers.RidgeStochasticGradientDescent,\n",
|
||||
" optimizers.RidgeMomentumSGD,\n",
|
||||
" optimizers.RidgeAdaGradSGD,\n",
|
||||
" optimizers.RidgeRMSPropSGD,\n",
|
||||
" optimizers.RidgeAdamSGD,\n",
|
||||
" ],\n",
|
||||
" optimizers.RIDGE_SGD_OPTIMIZERS,\n",
|
||||
" optimizer_kwargs,\n",
|
||||
" )\n",
|
||||
"]\n",
|
||||
"optimizers_lasso = [\n",
|
||||
" (opt, {**kwargs, **lasso_kwargs})\n",
|
||||
" for opt, kwargs in zip(\n",
|
||||
" [\n",
|
||||
" optimizers.LASSOStochasticGradientDescent,\n",
|
||||
" optimizers.LASSOMomentumSGD,\n",
|
||||
" optimizers.LASSOAdaGradSGD,\n",
|
||||
" optimizers.LASSORMSPropSGD,\n",
|
||||
" optimizers.LASSOAdamSGD,\n",
|
||||
" ],\n",
|
||||
" optimizers.LASSO_SGD_OPTIMIZERS,\n",
|
||||
" optimizer_kwargs,\n",
|
||||
" )\n",
|
||||
"]\n",
|
||||
"\n",
|
||||
"fig, axs = plt.subplots(1, 3, figsize=np.array(plotting.get_figsize(0.5)) * 2)\n",
|
||||
"plot_optimizers(\n",
|
||||
" axs[0], optimizers_ols, X_tr, y_tr, ylabel=\"Average Cost per Epoch (OLS)\"\n",
|
||||
"labels = [\"SGD\", \"Mom. SGD\", \"AdaGrad\", \"RMSProp\", \"Adam\"]\n",
|
||||
"\n",
|
||||
"fig, axs = plt.subplots(1, 3, figsize=plotting.get_figsize(0.5), sharey=True)\n",
|
||||
"plotting.plot_optimizers(\n",
|
||||
" axs[0], optimizers_ols, X_tr, y_tr, ylabel=\"Average Cost per Epoch\", labels=labels\n",
|
||||
")\n",
|
||||
"plot_optimizers(\n",
|
||||
" axs[1], optimizers_ridge, X_tr, y_tr, ylabel=\"Average Cost per Epoch (Ridge)\"\n",
|
||||
"plotting.plot_optimizers(\n",
|
||||
" axs[1], optimizers_ridge, X_tr, y_tr, ylabel=\"Average Cost per Epoch\", labels=labels\n",
|
||||
")\n",
|
||||
"plot_optimizers(\n",
|
||||
" axs[2], optimizers_lasso, X_tr, y_tr, ylabel=\"Average Cost per Epoch (LASSO)\"\n",
|
||||
"plotting.plot_optimizers(\n",
|
||||
" axs[2], optimizers_lasso, X_tr, y_tr, ylabel=\"Average Cost per Epoch\", labels=labels\n",
|
||||
")\n",
|
||||
"for ax in axs:\n",
|
||||
" ax.set_xlabel(\"Epoch\")\n",
|
||||
"fig.tight_layout()"
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"stochastic_gradient_descent_convergence.pdf\"))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "18",
|
||||
"id": "17",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"x = np.linspace(-1, 1, 300)\n",
|
||||
"y = datamanip.noise_data(datamanip.runge_function(x), 0.1)\n",
|
||||
"y = datamanip.noise_data(datamanip.runge_function(x), 1.0)\n",
|
||||
"x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2)\n",
|
||||
"\n",
|
||||
"polynomial_degrees = np.arange(1, 50)\n",
|
||||
@@ -676,7 +571,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "19",
|
||||
"id": "18",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -726,7 +621,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "20",
|
||||
"id": "19",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -750,7 +645,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "21",
|
||||
"id": "20",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -786,13 +681,14 @@
|
||||
"ax.set_ylabel(\"Mean Squared Error\")\n",
|
||||
"ax.set_yscale(\"log\")\n",
|
||||
"ax.legend()\n",
|
||||
"fig.tight_layout()"
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"k_fold_bootstrapping_comparision.pdf\"))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "22",
|
||||
"id": "21",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -830,7 +726,7 @@
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "23",
|
||||
"id": "22",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
@@ -856,16 +752,15 @@
|
||||
"\n",
|
||||
"ax.set_xlabel(\"Polynomial Degree\")\n",
|
||||
"ax.set_ylabel(f\"MSE ({k_folds}-fold validation)\")\n",
|
||||
"# ax.set_yscale(\"log\")\n",
|
||||
"# ax.set_ylim(0, 2)\n",
|
||||
"ax.legend()\n",
|
||||
"fig.tight_layout()"
|
||||
"fig.tight_layout()\n",
|
||||
"fig.savefig(os.path.join(FIG_DIR, \"kfold_mse_comparison_per_cost_function.pdf\"))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "24",
|
||||
"id": "23",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
|
||||
Reference in New Issue
Block a user