Skip to content

Instantly share code, notes, and snippets.

@calvinmccarter

calvinmccarter/bladderbatch-tabpfn.ipynb

Created November 26, 2023 22:10
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save calvinmccarter/8efdbcf39696462e5ec4b7a6fb9fd972 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save calvinmccarter/8efdbcf39696462e5ec4b7a6fb9fd972 to your computer and use it in GitHub Desktop.
Download ZIP
bladderbatch-tabpfn.ipynb
Display the source blob
Display the rendered blob
Raw
bladderbatch-tabpfn.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "1bef6bc4-aa17-44ec-823f-642fe8ba189a",
"metadata": {
"tags": []
},
"outputs": [],
"source": [
"import os\n",
"import warnings\n",
"import pandas as pd\n",
"import numpy as np\n",
"import seaborn as sns\n",
"\n",
"import sklearn as skl\n",
"import matplotlib.pyplot as plt\n",
"import sklearn.metrics as skmr\n",
"import xgboost as xgb\n",
"from tabpfn import TabPFNClassifier\n",
"from sklearn.svm import SVC\n",
"from sklearn.model_selection import train_test_split"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "1d80a572-6c4b-4d9f-8fbf-d03ff9738de3",
"metadata": {
"tags": []
},
"outputs": [],
"source": [
"# https://bioconductor.org/packages/release/data/experiment/html/bladderbatch.html\n",
"# bladder-expr.txt:\n",
"# https://drive.google.com/file/d/1Vq3xTJ3Tlm_NH8FnvnmeIonZy-J_iyK2/view?usp=sharing\n",
"# bladder-pheno.txt:\n",
"# https://drive.google.com/file/d/1VnpBFsNaHEWQalVXsAHtckluOCG6s1EH/view?usp=sharing\n",
"\n",
"pheno_orig = pd.read_table('bladder-pheno.txt', index_col=0)\n",
"expr_orig = pd.read_table('bladder-expr.txt', index_col=0).T"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "14eee78a-b95d-4aa7-8787-b940f46625a5",
"metadata": {
"collapsed": true,
"jupyter": {
"outputs_hidden": true
},
"tags": []
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Using a Transformer with 25.82 M parameters\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:09] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:18] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:26] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:34] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:42] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:50] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:04:58] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:05:06] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:05:14] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
"STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
"\n",
"Increase the number of iterations (max_iter) or scale the data as shown in:\n",
" https://scikit-learn.org/stable/modules/preprocessing.html\n",
"Please also refer to the documentation for alternative solver options:\n",
" https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
" extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
"/Users/calvinm/miniconda3/envs/tabpfn/lib/python3.7/site-packages/xgboost/sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
" warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[17:05:22] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
"WARNING: The number of features for this classifier is restricted to 100 and will be subsampled.\n"
]
}
],
"source": [
"methods = [\n",
" (\"SVC\", SVC()),\n",
" (\"LR\", skl.linear_model.LogisticRegression()),\n",
" (\"XGBoost\", xgb.XGBClassifier()),\n",
" (\"TabPFN\", TabPFNClassifier(device='cpu', subsample_features=True)),\n",
"]\n",
"n_random = 10\n",
"accs = np.zeros((len(methods), n_random))\n",
"f1s = np.zeros((len(methods), n_random))\n",
"for rix in range(n_random):\n",
" X_train, X_test, y_train, y_test = train_test_split(\n",
" expr_orig, pheno_orig.cancer, test_size=0.25, random_state=rix)\n",
" for mix, (mname, method) in enumerate(methods):\n",
" model = method\n",
" model.fit(X_train, y_train)\n",
" y_test_pred = model.predict(X_test)\n",
" accs[mix, rix] = skmr.accuracy_score(y_test, y_test_pred)\n",
" f1s[mix, rix] = skmr.f1_score(y_test, y_test_pred, average=\"macro\")"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "b78b580a-4e77-4ba5-980e-dfe844b86b98",
"metadata": {
"tags": []
},
"outputs": [
{
"data": {
"text/plain": [
"<Figure size 640x480 with 0 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 500x500 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"<Figure size 640x480 with 0 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 500x500 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"accs_df = pd.DataFrame(data=accs, index=[mname for mname, _ in methods]).T\n",
"f1s_df = pd.DataFrame(data=f1s, index=[mname for mname, _ in methods]).T\n",
"\n",
"plt.figure()\n",
"g = sns.catplot(data=accs_df, kind=\"violin\", inner=None)\n",
"sns.swarmplot(data=accs_df, color=\"k\", size=3, ax=g.ax)\n",
"plt.title(\"Accuracy on cancer\");\n",
"\n",
"plt.figure()\n",
"g = sns.catplot(data=f1s_df, kind=\"violin\", inner=None)\n",
"sns.swarmplot(data=f1s_df, color=\"k\", size=3, ax=g.ax)\n",
"plt.title(\"F1-macro on cancer\");"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"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": "3.7.16"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment