{ "cells": [ { "cell_type": "code", "execution_count": null, "id": "e9d321c1-26b8-43c9-80c6-fb1134394b14", "metadata": {}, "outputs": [], "source": [ "# ML prediction pipeline" ] }, { "cell_type": "markdown", "id": "e5b06b60-b1cf-4e87-bdd4-be7155fb6ac8", "metadata": {}, "source": [ "# Model development using different ML methods" ] }, { "cell_type": "raw", "id": "60557681-c58a-4a72-b7f6-159cab6c19bb", "metadata": {}, "source": [ "This script is to develop models using different ML methods and use nested 5-fold cross validation to validate the model performance. This pipeline does not include data preprocessing steps; please ensure you do data cleaning (e.g., nan, outliers) before using this pipeline\n", "\n", "Nested CV multipmodel training utility with threshold tuning and SHAP summaries.\n", "\n", "\n", "This module contains a compact collection of helper functions and a main routine\n", "(`my_nestedcv_multi`) to run a **nested cross‑validation** workflow across\n", "multiple classifiers using imbalanced-learn pipelines, with:\n", "\n", "\n", "- Standardization (pre‑SMOTE for meaningful distances)\n", "- BorderlineSMOTE oversampling\n", "- Model‑based feature selection\n", "- Grid search on inner CV (recall‑optimized)\n", "- Threshold tuning on the inner CV (Youden / F-beta / target sensitivity / cost)\n", "- Out‑of‑fold ROC aggregation\n", "- SHAP summary importances per feature" ] }, { "cell_type": "markdown", "id": "7b33e284-75cb-4aeb-b84b-1f9149d632e9", "metadata": {}, "source": [ "## Model training and validation functions" ] }, { "cell_type": "raw", "id": "0a172f13-632e-456e-b1cb-2730fe0b9b7c", "metadata": {}, "source": [ "main toolbox used\n", "# conda install -y -c conda-forge \\\n", " # \"numpy=1.26.4\" \\\n", " # \"scipy<1.14\" \\\n", " # \"pandas=2.2.*\" \\\n", " # \"scikit-learn=1.3.2\" \\\n", " # \"scikit-learn-intelex=2024.6.0\" \\\n", " # \"daal4py=2024.6.0\" \\\n", " # \"imbalanced-learn=0.13.0\" \\\n", " # \"shap=0.43.0\" \\\n", " # \"xgboost=1.7.6\" \\\n", " # \"tqdm: 4.67.1\"\n", " # matplotlib=3.8.4\n", " #\n" ] }, { "cell_type": "code", "execution_count": 23, "id": "e5cb3916-c747-4e03-88ce-95fa3d09f96a", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "Intel(R) Extension for Scikit-learn* enabled (https://github.com/intel/scikit-learn-intelex)\n" ] } ], "source": [ "# --- imports ---\n", "import os\n", "import time\n", "# Prevent numba from loading LLVM OpenMP\n", "os.environ[\"SHAP_DISABLE_JIT\"] = \"1\" # shap will use pure-python fallbacks\n", "os.environ[\"JOBLIB_PROGRESSBAR\"] = \"0\" # turn off joblib's tqdm bars\n", "\n", "# Standard library\n", "import statistics\n", "\n", "# Core scientific stack\n", "import numpy as np\n", "import pandas as pd\n", "\n", "# Utilities\n", "from tqdm import tqdm # To display progress bars during loops and long computations.\n", "\n", "# # Scikit-learn with Intel acceleration\n", "from sklearnex import patch_sklearn\n", "patch_sklearn()\n", "\n", "# Scikit-learn modules\n", "from sklearn.decomposition import PCA\n", "from sklearn.model_selection import StratifiedKFold, GridSearchCV, cross_val_predict\n", "from sklearn.metrics import accuracy_score, confusion_matrix, roc_curve, auc, f1_score\n", "from sklearn.preprocessing import StandardScaler\n", "from sklearn.feature_selection import SelectFromModel\n", "\n", "\n", "# Classifiers\n", "from sklearn.ensemble import RandomForestClassifier\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.svm import SVC\n", "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.neighbors import KNeighborsClassifier\n", "\n", "# Imbalanced-learn\n", "from imblearn.pipeline import Pipeline as ImbPipeline\n", "from imblearn.over_sampling import BorderlineSMOTE, SMOTE\n", "\n", "# Model interpretation\n", "import shap\n", "\n", "# Gradient boosting\n", "from xgboost import XGBClassifier\n", "\n", "\n", "\n", "def _gpu_available():\n", " \"\"\"Return True if a CUDA-capable GPU appears available via numba.\n", " \n", " \n", " Notes\n", " -----\n", " - This is a lightweight check. XGBoost may still error if CUDA runtime is\n", " unavailable or mismatched even when this returns True.\n", " \"\"\"\n", " try:\n", " import numba.cuda as _cuda\n", " return _cuda.is_available()\n", " except Exception:\n", " return False\n", " \n", "def arraylist2arrayColumn(arrays):\n", " \"\"\"Stack a list of 1‑D arrays (possibly different lengths) as columns.\n", " \n", " \n", " Shorter arrays are **left‑padded with zeros** so that all columns align.\n", " \n", " \n", " Parameters\n", " ----------\n", " arrays : list[np.ndarray]\n", " List of 1‑D arrays with potentially different lengths.\n", " \n", " \n", " Returns\n", " -------\n", " np.ndarray\n", " 2‑D array of shape (max_length, n_arrays) where each input array forms\n", " a column and is left‑padded with zeros to `max_length`.\n", " \"\"\"\n", " max_length = max(len(arr) for arr in arrays)\n", " padded_arrays = [np.pad(arr, (max_length - len(arr), 0), constant_values=0) for arr in arrays]\n", " return np.column_stack(padded_arrays)\n", "\n", "def allocate_feature_list(data, feature_list):\n", " \"\"\"Count how often features were selected across folds/iterations.\n", " \n", " \n", " Parameters\n", " ----------\n", " data : pd.DataFrame\n", " Original feature dataframe; used to preserve full feature list/order.\n", " feature_list : list[Iterable[str]]\n", " List where each item is an iterable of selected feature names from one\n", " fold/iteration.\n", " \n", " \n", " Returns\n", " -------\n", " pd.DataFrame\n", " DataFrame with columns `Feature` and `Count` (selection frequency).\n", " \"\"\"\n", " all_features = list(data.columns)\n", " flattened = [f for sub in feature_list for f in sub]\n", " counts = pd.Series(flattened).value_counts()\n", " out = pd.DataFrame({'Feature': all_features})\n", " out['Count'] = out['Feature'].map(counts).fillna(0).astype(int)\n", " return out\n", "\n", "def should_apply_pca(x, threshold=1.5):\n", " \"\"\"Heuristic to decide if PCA can be applied based on sample/feature ratio. if false, pca is not allowed.\n", " \n", " Parameters\n", " ----------\n", " x : pd.DataFrame | np.ndarray\n", " Feature matrix.\n", " threshold : float, default 1.5\n", " Apply PCA when n_samples / n_features < `threshold`.\n", " \n", " \n", " Returns\n", " -------\n", " bool\n", " True if PCA is recommended under the heuristic.\n", " \"\"\"\n", " n_samples, n_features = x.shape\n", " return n_samples / n_features < threshold\n", "\n", "\n", "def _to_2d_array(sv):\n", " \"\"\"Normalize SHAP output to shape (n_samples, n_features).\n", " \n", " \n", " Handles lists (per‑class), 1D vectors, and 3D arrays.\n", " \n", " \n", " Parameters\n", " ----------\n", " sv : array-like | list\n", " Raw SHAP values as returned by various explainers/models.\n", " \n", " \n", " Returns\n", " -------\n", " np.ndarray\n", " SHAP values as a 2‑D array.\n", " \"\"\"\n", " # If SHAP returns a list per class -> take positive class if available\n", " if isinstance(sv, (list, tuple)):\n", " sv = sv[1] if len(sv) > 1 else sv[0]\n", " sv = np.asarray(sv)\n", "\n", " # If 3D: (n_samples, n_features, n_outputs). Pick class 1 if available; else 0.\n", " if sv.ndim == 3:\n", " sv = sv[..., 1] if sv.shape[2] > 1 else sv[..., 0]\n", "\n", " # If 1D: (n_features,) -> (1, n_features)\n", " if sv.ndim == 1:\n", " sv = sv.reshape(1, -1)\n", "\n", " # Now it should be (n_samples, n_features)\n", " return sv\n", "\n", "\n", "\n", "# ---------- model specs ----------\n", "def _get_model_specs(use_gpu=False, n_jobs=5, ratio=1.0):\n", " \"\"\"Define available model, grid, selector, and type for the pipeline search.\n", " \n", " \n", " Parameters\n", " ----------\n", " use_gpu : bool, default False\n", " If True, enable GPU‑accelerated XGBoost when available.\n", " n_jobs : int, default 5\n", " Parallelism for estimators that support it.\n", " ratio : float, default 1.0\n", " Passed to XGBoost as `scale_pos_weight` to compensate class imbalance\n", " (usually `neg/pos`).\n", " \n", " \n", " Returns\n", " -------\n", " dict[str, dict]\n", " Mapping from model name to its specification with keys:\n", " - \"clf\": estimator instance\n", " - \"grid\": parameter grid for GridSearchCV\n", " - \"selector\": feature selector (SelectFromModel)\n", " - \"type\": one of {\"linear\", \"kernel\", \"tree\"}\n", " \"\"\"\n", " rf_selector = SelectFromModel(RandomForestClassifier(n_estimators=200, random_state=42), threshold='median')\n", "\n", " specs = {\n", " # lasso and ridge are just penalties for logistic, so I comment them here.\n", " # \"lasso\": {\n", " # \"clf\": LogisticRegression(penalty='l1', solver='saga', max_iter=5000),\n", " # \"grid\": {\n", " # \"feature_selection__threshold\": ['mean', 'median'],\n", " # \"classification__C\": [0.01, 0.1, 1.0, 10.0],\n", " # },\n", " # \"selector\": SelectFromModel(LogisticRegression(penalty='l1', solver='saga', max_iter=5000), threshold='median'),\n", " # \"type\": \"linear\"\n", " # },\n", " # \"ridge\": {\n", " # \"clf\": LogisticRegression(penalty='l2', solver='lbfgs', max_iter=5000),\n", " # \"grid\": {\n", " # \"feature_selection__threshold\": ['mean', 'median'],\n", " # \"classification__C\": [0.01, 0.1, 1.0, 10.0],\n", " # },\n", " # \"selector\": SelectFromModel(LogisticRegression(penalty='l2', solver='lbfgs', max_iter=5000), threshold='median'),\n", " # \"type\": \"linear\"\n", " # },\n", " \"logistic\": {\n", " \"clf\": LogisticRegression(class_weight='balanced', penalty='l2', solver='lbfgs', max_iter=5000),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__C\": [0.01, 0.1, 1.0, 10.0],\n", " },\n", " \"selector\": SelectFromModel(LogisticRegression(penalty='l2', solver='lbfgs', max_iter=5000), threshold='median'),\n", " \"type\": \"linear\"\n", " },\n", " \"svm\": {\n", " \"clf\": SVC(kernel='rbf', class_weight= 'balanced' , probability=True),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__C\": [0.1, 1],\n", " \"classification__gamma\": ['auto']\n", " },\n", " \"selector\": rf_selector,\n", " \"type\": \"kernel\"\n", " },\n", " \"random_forest\": {\n", " \"clf\": RandomForestClassifier(class_weight='balanced', n_jobs=n_jobs, random_state=42),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__n_estimators\": [100, 200, 400],\n", " \"classification__min_samples_leaf\": [1, 2],\n", " \"classification__max_features\": ['sqrt', 0.5]\n", " },\n", " \"selector\": SelectFromModel(RandomForestClassifier(n_estimators=200, random_state=42), threshold='median'),\n", " \"type\": \"tree\"\n", " },\n", " \"decision_tree\": {\n", " \"clf\": DecisionTreeClassifier(class_weight='balanced', random_state=42),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__max_depth\": [5, 10],\n", " \"classification__min_samples_split\": [2, 5]\n", " },\n", " \"selector\": rf_selector,\n", " \"type\": \"tree\"\n", " },\n", " \"knn\": {\n", " \"clf\": KNeighborsClassifier(),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__n_neighbors\": [3, 5, 7, 11],\n", " \"classification__weights\": ['uniform', 'distance'],\n", " \"classification__p\": [1, 2],\n", " },\n", " \"selector\": rf_selector,\n", " \"type\": \"kernel\"\n", " },\n", " }\n", "\n", "\n", " xgb_kwargs = dict(\n", " n_estimators=300,\n", " learning_rate=0.05,\n", " subsample=0.9,\n", " colsample_bytree=0.9,\n", " max_depth=4,\n", " reg_lambda=1.0,\n", " reg_alpha=0.0,\n", " eval_metric='logloss',\n", " scale_pos_weight=ratio,\n", " n_jobs=n_jobs,\n", " random_state=42,\n", " # use_label_encoder=False\n", " )\n", " if use_gpu and _gpu_available():\n", " xgb_kwargs.update({\n", " \"tree_method\": \"gpu_hist\",\n", " \"predictor\": \"gpu_predictor\"\n", " })\n", " # xgb_kwargs.update({\n", " # \"tree_method\": \"hist\",\n", " # \"predictor\": \"auto\",\n", " # \"device\": \"cuda\"\n", " # })\n", " else:\n", " # fast CPU histogram as fallback\n", " xgb_kwargs.update({\n", " \"tree_method\": \"hist\",\n", " # \"predictor\": \"auto\"\n", " })\n", "\n", " specs[\"xgboost\"] = {\n", " \"clf\": XGBClassifier(**xgb_kwargs),\n", " \"grid\": {\n", " \"feature_selection__threshold\": ['median'],\n", " \"classification__n_estimators\": [200, 400],\n", " \"classification__max_depth\": [4, 6],\n", " \"classification__learning_rate\": [0.03, 0.1],\n", " \"classification__subsample\": [0.8],\n", " \"classification__colsample_bytree\": [0.8],\n", " },\n", " \"selector\": SelectFromModel(RandomForestClassifier(n_estimators=200, random_state=42), threshold='median'),\n", " \"type\": \"tree\"\n", " }\n", " return specs\n", "\n", "\n", "# ---------- SHAP utils ----------\n", "def _compute_shap_summary(best_pipeline, x_train, x_test, feature_names, model_type,\n", " bg_max=100, test_max=300):\n", " \"\"\"\n", " Compute mean|SHAP| per (selected) feature for the final estimator in best_pipeline.\n", " Returns a DataFrame with columns: Feature, mean_abs_shap.\n", " For SVM/KNN (kernel), uses KernelExplainer with small background/test samples.\n", " \"\"\"\n", "\n", " \"\"\"Compute mean |SHAP| per (selected) feature for the pipeline's classifier.\n", " \n", " \n", " Parameters\n", " ----------\n", " best_pipeline : imblearn.Pipeline\n", " Fitted pipeline with steps [scaler, smote, feature_selection, classification].\n", " x_train, x_test : pd.DataFrame | np.ndarray\n", " Train/test data from the outer fold (before scaling/selection).\n", " feature_names : list[str]\n", " Original feature names for alignment.\n", " model_type : {\"tree\", \"linear\", \"kernel\"}\n", " Determines which SHAP explainer to use.\n", " bg_max : int, default 100\n", " Background sample size for KernelExplainer (used for kernel models).\n", " test_max : int, default 300\n", " Test subsample size for KernelExplainer (used for kernel models).\n", " \n", " \n", " Returns\n", " -------\n", " pd.DataFrame\n", " Columns: `Feature`, `mean_abs_shap` for **all** original features\n", " (unselected features are filled with 0.0). If SHAP fails, values are NaN.\n", " \n", " \n", " Notes\n", " -----\n", " - For kernel models (SVM/KNN), full SHAP on all samples is expensive. We\n", " approximate by computing SHAP on a small subset and taking the mean |SHAP|.\n", " - For tree/linear models, TreeExplainer/LinearExplainer is used directly.\n", " \"\"\"\n", "\n", "\n", " # Extract steps\n", " scaler = best_pipeline.named_steps.get('scaler', None)\n", " selector = best_pipeline.named_steps['feature_selection']\n", " est = best_pipeline.named_steps['classification']\n", "\n", " # Transform to the estimator's input space (selected features after scaling)\n", " # Xtr = x_train.values if isinstance(x_train, pd.DataFrame) else x_train\n", " # Xte = x_test.values if isinstance(x_test, pd.DataFrame) else x_test\n", " Xtr = x_train if isinstance(x_train, pd.DataFrame) else pd.DataFrame(x_train, columns=feature_names)\n", " Xte = x_test if isinstance(x_test, pd.DataFrame) else pd.DataFrame(x_test, columns=feature_names)\n", "\n", " if scaler is not None:\n", " Xtr = scaler.transform(Xtr)\n", " Xte = scaler.transform(Xte)\n", "\n", " Xtr_sel = selector.transform(Xtr)\n", " Xte_sel = selector.transform(Xte)\n", "\n", " # Map selected column names\n", " support = selector.get_support()\n", " # tqdm.write(f\"[{model_type}] selected features: {np.sum(support)} / {len(support)}\")\n", " if support is None or np.sum(support) == 0:\n", " sel_names = np.array(feature_names)\n", " else:\n", " sel_names = np.array(feature_names)[support]\n", "\n", " # If no features selected (edge case), return NaNs\n", " if Xtr_sel.shape[1] == 0:\n", " return pd.DataFrame({\"Feature\": feature_names, \"mean_abs_shap\": np.nan})\n", "\n", " # Choose explainer\n", " try:\n", " if model_type == \"tree\":\n", " explainer = shap.TreeExplainer(est)\n", " shap_vals = explainer.shap_values(Xte_sel)\n", " # binary: list-of-arrays per class; take class 1 if available\n", " if isinstance(shap_vals, list):\n", " # prefer positive class if present\n", " sv = shap_vals[1] if len(shap_vals) > 1 else shap_vals[0]\n", " else:\n", " sv = shap_vals\n", " \n", " elif model_type == \"linear\":\n", " explainer = shap.LinearExplainer(est, Xtr_sel)\n", " sv = explainer.shap_values(Xte_sel)\n", " \n", " else: # \"kernel\" (e.g., SVM, KNN)\n", " # sample small background and test to keep it fast\n", " rng = np.random.default_rng(0)\n", " bg_idx = rng.choice(Xtr_sel.shape[0], size=min(bg_max, Xtr_sel.shape[0]), replace=False)\n", " te_idx = rng.choice(Xte_sel.shape[0], size=min(test_max, Xte_sel.shape[0]), replace=False)\n", " Xbg = Xtr_sel[bg_idx]\n", " Xte_small = Xte_sel[te_idx]\n", " # explain probability of class 1\n", " f = (lambda A: est.predict_proba(A)[:, 1])\n", " explainer = shap.KernelExplainer(f, Xbg) # Kernel SHAP approximates this by replacing missing features with their distribution from the background dataset.\n", " sv_small = explainer.shap_values(Xte_small, nsamples=\"auto\", silent=True)\n", " # expand to full test by mapping means (approx)\n", " sv = np.zeros((Xte_sel.shape[0], Xte_sel.shape[1]))\n", " sv[:] = np.mean(np.abs(sv_small), axis=0) # use mean abs across the small set as proxy\n", " # NOTE: Kernel SHAP exact per-sample is expensive; this proxy summarizes feature impact.\n", " sv = _to_2d_array(sv)\n", " mean_abs = np.mean(np.abs(sv), axis=0) if sv.ndim == 2 else np.mean(np.abs(sv), axis=0)\n", " df = pd.DataFrame({\"Feature\": sel_names, \"mean_abs_shap\": mean_abs})\n", " # ensure all features appear; fill 0 for unselected\n", " out = pd.DataFrame({\"Feature\": feature_names}).merge(df, on=\"Feature\", how=\"left\").fillna(0.0)\n", " return out\n", "\n", " except Exception:\n", " # fallback if SHAP fails\n", " return pd.DataFrame({\"Feature\": feature_names, \"mean_abs_shap\": np.nan})\n", "\n", "\n", "\n", "def find_threshold(y_true, p_val, mode=\"fbeta\", beta=2.0, target_sens=None, costs=None):\n", " \"\"\"\n", " Pick an operating threshold from validation probs.\n", " modes:\n", " - \"fbeta\": maximize F_beta (beta>1 favors sensitivity)\n", " - \"youden\": maximize Youden's J = TPR - FPR\n", " - \"target_sens\": smallest threshold with sensitivity >= target_sens\n", " - \"cost\": minimize expected cost given costs={\"fp\":c_fp,\"fn\":c_fn}\n", " beta : float, default 2.0\n", " F-beta parameter (used only when mode=\"fbeta\").\n", " target_sens : float | None\n", " Target sensitivity (recall) in [0,1] when mode=\"target_sens\".\n", " costs : dict | None\n", " Required when mode=\"cost\"; keys {\"fp\", \"fn\"} specify relative costs for\n", " false positives/negatives.\n", " \n", " \n", " Returns\n", " -------\n", " float\n", " Chosen threshold in [0,1].\n", " \"\"\"\n", " # use all unique prob cutpoints (plus 0 and 1 for safety)\n", " thr_grid = np.unique(np.r_[0.0, p_val, 1.0])\n", " best_thr, best_val = 0.5, -np.inf\n", " if mode == \"cost\": \n", " # for costs we minimize, so track min instead\n", " best_val, best_thr = np.inf, 0.5\n", "\n", " for t in thr_grid:\n", " y_hat = (p_val >= t).astype(int)\n", " tp = np.sum((y_true==1) & (y_hat==1))\n", " fn = np.sum((y_true==1) & (y_hat==0))\n", " tn = np.sum((y_true==0) & (y_hat==0))\n", " fp = np.sum((y_true==0) & (y_hat==1))\n", " sens = tp / (tp + fn) if (tp+fn)>0 else 0.0\n", " spec = tn / (tn + fp) if (tn+fp)>0 else 0.0\n", "\n", " if mode == \"fbeta\":\n", " prec = tp / (tp + fp) if (tp+fp)>0 else 0.0\n", " rec = sens\n", " if prec+rec==0:\n", " val = 0.0\n", " else:\n", " beta2 = beta*beta\n", " val = (1+beta2) * (prec*rec) / (beta2*prec + rec) if (beta2*prec + rec)>0 else 0.0\n", " if val > best_val:\n", " best_val, best_thr = val, t\n", "\n", " elif mode == \"youden\":\n", " val = sens - (1 - spec) # TPR - FPR\n", " if val > best_val:\n", " best_val, best_thr = val, t\n", "\n", " elif mode == \"target_sens\":\n", " # choose smallest t achieving desired sensitivity\n", " if target_sens is None:\n", " raise ValueError(\"Provide target_sens for mode='target_sens'\")\n", " if sens >= target_sens:\n", " # prefer higher specificity among those meeting the target; break ties by higher threshold\n", " val = spec\n", " if val > best_val or (np.isclose(val, best_val) and t > best_thr):\n", " best_val, best_thr = val, t\n", "\n", " elif mode == \"cost\":\n", " if costs is None or not {\"fp\",\"fn\"} <= set(costs):\n", " raise ValueError(\"Provide costs={'fp':...,'fn':...} for mode='cost'\")\n", " exp_cost = costs[\"fp\"]*fp + costs[\"fn\"]*fn\n", " if exp_cost < best_val:\n", " best_val, best_thr = exp_cost, t\n", "\n", " else:\n", " raise ValueError(\"Unknown mode\")\n", "\n", " return float(best_thr)\n", " \n", "\n", "def tune_threshold_on_inner(x_train, y_train, best_pipeline, inner_cv,\n", " mode=\"fbeta\", beta=2.0, target_sens=None, costs=None, n_jobs=1):\n", " \"\"\"\n", " Using the best pipeline (hyperparams fixed), get out-of-fold validation probabilities\n", " on x_train via the inner CV, then pick threshold by the chosen rule.\n", " Parameters\n", " ----------\n", " x_train : pd.DataFrame | np.ndarray\n", " Features for the outer fold's training split.\n", " y_train : pd.Series | np.ndarray\n", " Binary labels for the outer fold's training split.\n", " best_pipeline : imblearn.Pipeline\n", " Pipeline with hyperparameters fixed (e.g., the GridSearchCV best estimator).\n", " inner_cv : sklearn.model_selection.StratifiedKFold\n", " The same inner CV scheme used during hyperparameter selection.\n", " mode, beta, target_sens, costs : see `find_threshold`.\n", " n_jobs : int, default 1\n", " Parallelism for `cross_val_predict`.\n", " \n", " \n", " Returns\n", " -------\n", " float\n", " Selected decision threshold.\n", " \"\"\"\n", " # get OOF probabilities for the positive class\n", " p_val = cross_val_predict(best_pipeline, x_train, y_train,\n", " cv=inner_cv, method=\"predict_proba\", n_jobs=n_jobs)[:, 1]\n", " thr = find_threshold(y_train.values, p_val,\n", " mode=mode, beta=beta, target_sens=target_sens, costs=costs)\n", " return thr\n", "\n", "\n", "\n", "\n", "\n", "\n", "# ---------- main multi-model nested CV with tqdm, scaler, SHAP ----------\n", "def my_nestedcv_multi(\n", " x, y, n_iter=1, n_jobs=5, permutation=False, threshold_mode=\"youden\", pca_predictors=False,\n", " model_names=(\"svm\", \"logistic\", \"random_forest\", \"decision_tree\", \"xgboost\", \"knn\"),\n", " shap_kernel_bg_max=30, shap_kernel_test_max=40, use_gpu=True, ratio=1.0\n", "):\n", " \"\"\"Run nested CV for multiple models with SMOTE, feature selection, and SHAP.\n", " \n", " \n", " Workflow (per model)\n", " --------------------\n", " For each outer fold and `n_iter` repetitions:\n", " 1) Inner 5× Stratified CV grid search (scoring='recall') over pipeline params\n", " 2) Threshold tuning on inner CV OOF probabilities (e.g., Youden)\n", " 3) Refit best pipeline on full outer-train with tuned threshold\n", " 4) Evaluate on outer-test (ACC, SEN, SPEC, PPV, NPV, AUC + ROC curve)\n", " 5) Record selected features and SHAP mean |SHAP| per feature\n", " \n", " \n", " Parameters\n", " ----------\n", " x : pd.DataFrame\n", " Feature matrix. If `pca_predictors=True` and the heuristic triggers,\n", " PCA is applied once to reduce dimensionality.\n", " y : array-like\n", " Binary labels (0/1).\n", " n_iter : int, default 1\n", " How many times to repeat the outer CV (with different shuffles).\n", " n_jobs : int, default 5\n", " Parallelism for GridSearchCV and models that support it.\n", " permutation : bool, default False\n", " If True, permutes labels for a quick sanity check (should degrade performance).\n", " threshold_mode : str, default \"youden\"\n", " Passed to `tune_threshold_on_inner` (\"youden\", \"fbeta\", \"target_sens\", \"cost\").\n", " pca_predictors : bool, default False\n", " Whether to optionally apply PCA before CV when samples/features are low.\n", " model_names : tuple[str], default (\"svm\", \"logistic\", \"random_forest\", \"decision_tree\", \"xgboost\", \"knn\")\n", " Subset of models defined in `_get_model_specs` to run.\n", " shap_kernel_bg_max, shap_kernel_test_max : int\n", " Subsample sizes for Kernel SHAP when using kernel models (SVM/KNN).\n", " use_gpu : bool, default True\n", " Try GPU for XGBoost if available.\n", " ratio : float, default 1.0\n", " Class imbalance ratio for XGBoost's `scale_pos_weight`.\n", " \n", " \n", " Returns\n", " -------\n", " dict\n", " Mapping: model_name -> {\n", " \"performance\": {\n", " \"score\": {\"accuracy\": list, \"sensitivity\": list, \"specificity\": list,\n", " \"AUC\": list, \"PPV\": list, \"NPV\": list},\n", " \"fpr\": list[np.ndarray],\n", " \"tpr\": list[np.ndarray],\n", " },\n", " \"selected_features\": pd.DataFrame, # selection counts across folds/iters\n", " \"shap_summary\": pd.DataFrame # mean of mean|SHAP| across folds/iters\n", " }\n", " \n", " \n", " Notes\n", " -----\n", " - The pipeline order is critical: **StandardScaler -> SMOTE -> FeatureSelection -> Classifier**.\n", " This ensures SMOTE operates in a standardized space and selection is applied to\n", " the resampled data.\n", " - GridSearchCV uses `scoring='recall'` to prioritize sensitivity on imbalanced data.\n", " \"\"\"\n", " # PCA (optional; applied once pre-CV ) would be better to put this into the \"ImbPipeline\" to avoid data leakage\n", " if pca_predictors:\n", " if should_apply_pca(x):\n", " n_components = min(10, x.shape[1])\n", " pca = PCA(n_components=n_components)\n", " x = pd.DataFrame(pca.fit_transform(x), index=x.index, columns=[f\"PC{i+1}\" for i in range(n_components)])\n", " else:\n", " x = x.copy()\n", " else:\n", " x = x.copy()\n", "\n", " # y to Series 0/1\n", " y_series = pd.Series(y).astype(int).reset_index(drop=True)\n", " if permutation:\n", " y_series = y_series.sample(frac=1.0, random_state=None).reset_index(drop=True)\n", "\n", " model_specs = _get_model_specs(use_gpu=use_gpu, n_jobs=n_jobs, ratio=ratio)\n", " results_by_model = {}\n", "\n", " outer_cv = StratifiedKFold(n_splits=5, shuffle=True)\n", " inner_cv = StratifiedKFold(n_splits=5, shuffle=True)\n", "\n", " feature_names = list(x.columns)\n", "\n", " # Only run models that exist in the spec mapping\n", " pbar_models = [m for m in model_names if m in model_specs]\n", " for name in pbar_models:\n", " spec = model_specs[name]\n", " classifier = spec[\"clf\"]\n", " param_grid = spec[\"grid\"]\n", " feature_selector = spec[\"selector\"]\n", " model_type = spec[\"type\"]\n", "\n", " \n", " # Build pipeline\n", " # NOTE: StandardScaler is placed **before** SMOTE so that k-NN based resamplers\n", " # and distance metrics are computed in a standardized space.\n", " pipeline = ImbPipeline([\n", " ('scaler', StandardScaler(with_mean=True, with_std=True)),\n", " ('smote', BorderlineSMOTE(kind='borderline-1', random_state=42)),\n", " # ('smote', SMOTE(random_state=42)),\n", " ('feature_selection', feature_selector),\n", " ('classification', classifier),\n", " ])\n", "\n", " overall_outercv_performance = {\n", " \"score\": {\"accuracy\": [], \"sensitivity\": [], \"specificity\": [], \"AUC\": [], \"PPV\": [], \"NPV\": []},\n", " \"fpr\": [],\n", " \"tpr\": [],\n", " }\n", " selected_features_list = []\n", " shap_summaries = [] # collect per-fold mean|SHAP|; aggregate later\n", "\n", " pbar_iters = tqdm(range(n_iter), desc=f\"{name} • iterations\", leave=True)\n", " for _ in pbar_iters:\n", " iteration_outercv = {\n", " \"score\": {\"accuracy\": [], \"sensitivity\": [], \"specificity\": [], \"AUC\": [], \"PPV\": [], \"NPV\": []},\n", " \"fpr\": [],\n", " \"tpr\": [],\n", " }\n", "\n", " # pbar_folds = tqdm(list(outer_cv.split(x, y_series)), desc=f\"{name} • outer folds\", leave=False)\n", " for train_idx, test_idx in outer_cv.split(x, y_series):\n", " x_train, x_test = x.iloc[train_idx], x.iloc[test_idx]\n", " y_train, y_test = y_series.iloc[train_idx], y_series.iloc[test_idx]\n", "\n", " # Inner CV grid search (optimize recall) you can choose others like roc_auc if needed\n", " grid_search = GridSearchCV(\n", " estimator=pipeline,\n", " param_grid=param_grid,\n", " cv=inner_cv,\n", " scoring='recall',\n", " n_jobs=n_jobs,\n", " verbose=0 # Silence joblib progress display\n", " )\n", " grid_search.fit(x_train, y_train)\n", " best_pipeline = grid_search.best_estimator_\n", "\n", "\n", " ## choose the best threshold\n", " best_thr = tune_threshold_on_inner(\n", " x_train, y_train, best_pipeline, inner_cv,\n", " mode=threshold_mode, beta=2.0, # e.g., favor sensitivity\n", " # OR: mode=\"youden\"\n", " # OR: mode=\"target_sens\", target_sens=0.80\n", " # OR: mode=\"cost\", costs={\"fp\":1, \"fn\":5} # cost of FN is higher \n", " n_jobs=n_jobs)\n", " \n", " # 2) refit on full outer-train (already fit by gridsearch on inner splits,\n", " # but we want a fresh fit on all x_train with the chosen params)\n", " best_pipeline.fit(x_train, y_train)\n", "\n", " # Selected features (names)\n", " selector = best_pipeline.named_steps['feature_selection']\n", " support = selector.get_support()\n", " if support is None or np.sum(support) == 0:\n", " selected_cols = x.columns\n", " else:\n", " selected_cols = x.columns[support]\n", " selected_features_list.append(selected_cols)\n", "\n", " # Evaluate on outer test with tuned threshold\n", " probs = best_pipeline.predict_proba(x_test)[:, 1]\n", " yhat = (probs >= best_thr).astype(int)\n", "\n", " acc = accuracy_score(y_test, yhat)\n", " tn, fp, fn, tp = confusion_matrix(y_test, yhat).ravel()\n", " sens = tp / (tp + fn) if (tp + fn) > 0 else 0.0\n", " spec = tn / (tn + fp) if (tn + fp) > 0 else 0.0\n", " cur_fpr, cur_tpr, _ = roc_curve(y_test, probs)\n", " ppv = tp / (tp + fp) if (tp + fp) > 0 else 0.0\n", " npv = tn / (tn + fn) if (tn + fn) > 0 else 0.0\n", " auc_val = auc(cur_fpr, cur_tpr)\n", "\n", " iteration_outercv[\"score\"][\"accuracy\"].append(acc)\n", " iteration_outercv[\"score\"][\"sensitivity\"].append(sens)\n", " iteration_outercv[\"score\"][\"specificity\"].append(spec)\n", " iteration_outercv[\"score\"][\"AUC\"].append(auc_val)\n", " iteration_outercv[\"score\"][\"PPV\"].append(ppv)\n", " iteration_outercv[\"score\"][\"NPV\"].append(npv)\n", " iteration_outercv[\"fpr\"].append(cur_fpr)\n", " iteration_outercv[\"tpr\"].append(cur_tpr)\n", "\n", " # SHAP summary for this fold (mean|SHAP| per feature)\n", " shap_df = _compute_shap_summary(\n", " best_pipeline, x_train, x_test, feature_names, model_type,\n", " bg_max=shap_kernel_bg_max, test_max=shap_kernel_test_max\n", " )\n", " shap_summaries.append(shap_df)\n", "\n", " # Aggregate metrics across outer folds for this iteration\n", " overall_outercv_performance[\"score\"][\"accuracy\"].append(statistics.mean(iteration_outercv[\"score\"][\"accuracy\"]))\n", " overall_outercv_performance[\"score\"][\"sensitivity\"].append(statistics.mean(iteration_outercv[\"score\"][\"sensitivity\"]))\n", " overall_outercv_performance[\"score\"][\"specificity\"].append(statistics.mean(iteration_outercv[\"score\"][\"specificity\"]))\n", " overall_outercv_performance[\"score\"][\"AUC\"].append(statistics.mean(iteration_outercv[\"score\"][\"AUC\"]))\n", " overall_outercv_performance[\"score\"][\"PPV\"].append(statistics.mean(iteration_outercv[\"score\"][\"PPV\"]))\n", " overall_outercv_performance[\"score\"][\"NPV\"].append(statistics.mean(iteration_outercv[\"score\"][\"NPV\"]))\n", " overall_outercv_performance[\"fpr\"].append(np.mean(arraylist2arrayColumn(iteration_outercv[\"fpr\"]), axis=1))\n", " overall_outercv_performance[\"tpr\"].append(np.mean(arraylist2arrayColumn(iteration_outercv[\"tpr\"]), axis=1))\n", "\n", " # Feature selection frequencies across all folds/iterations\n", " selected_features_df = allocate_feature_list(x, selected_features_list)\n", "\n", " # aggregate SHAP summaries across all folds/iterations (mean of mean|SHAP|)\n", " if len(shap_summaries) > 0:\n", " shap_concat = pd.concat(shap_summaries, axis=0)\n", " shap_summary = (shap_concat.groupby(\"Feature\", as_index=False)[\"mean_abs_shap\"]\n", " .mean()\n", " .sort_values(\"mean_abs_shap\", ascending=False)\n", " .reset_index(drop=True))\n", " else:\n", " shap_summary = pd.DataFrame({\"Feature\": feature_names, \"mean_abs_shap\": np.nan})\n", "\n", " results_by_model[name] = {\n", " \"performance\": overall_outercv_performance,\n", " \"selected_features\": selected_features_df,\n", " \"shap_summary\": shap_summary\n", " }\n", "\n", " return results_by_model\n" ] }, { "cell_type": "markdown", "id": "fa3567a3-6ee6-4414-ae1d-2081fe7302dc", "metadata": {}, "source": [ "## Load data" ] }, { "cell_type": "code", "execution_count": 24, "id": "aa5e29f7-0345-4e48-99e7-86db642c4ed5", "metadata": {}, "outputs": [], "source": [ "# create predictors and outcome\n", "# Load sheets\n", "data = pd.read_excel(r\"MLpipeline data.xlsx\")\n", "\n", "predictors = data.drop(columns=[\"outcome\"])\n", "\n", "# Create the outcome variable\n", "y = (data[\"outcome\"]).astype(int)\n", "\n", "y_ = np.asarray(y)\n", "n_pos = np.sum(y_ == 1)\n", "n_neg = np.sum(y_ == 0)\n", "ratio = float(n_neg) / float(max(n_pos, 1))\n", "\n", "n_iter = 5 # number of iteration\n", "n_jobs = 6 # number of cores in parallel computation" ] }, { "cell_type": "markdown", "id": "a090593d-a53d-4092-8266-7a37e9cdea87", "metadata": {}, "source": [ "## Run the ML pipeline" ] }, { "cell_type": "code", "execution_count": 25, "id": "26c3700f-a7b2-49ef-a572-d08d37ee1026", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "svm • iterations: 100%|██████████████████████████████████████████████████████████████████| 5/5 [03:00<00:00, 36.01s/it]\n", "logistic • iterations: 100%|█████████████████████████████████████████████████████████████| 5/5 [00:09<00:00, 1.87s/it]\n", "random_forest • iterations: 100%|████████████████████████████████████████████████████████| 5/5 [01:16<00:00, 15.32s/it]\n", "decision_tree • iterations: 100%|████████████████████████████████████████████████████████| 5/5 [00:39<00:00, 7.97s/it]\n", "xgboost • iterations: 100%|██████████████████████████████████████████████████████████████| 5/5 [01:43<00:00, 20.69s/it]\n", "knn • iterations: 100%|██████████████████████████████████████████████████████████████████| 5/5 [04:02<00:00, 48.59s/it]\n", "svm • iterations: 100%|██████████████████████████████████████████████████████████████████| 5/5 [03:00<00:00, 36.02s/it]\n", "logistic • iterations: 100%|█████████████████████████████████████████████████████████████| 5/5 [00:09<00:00, 1.85s/it]\n", "random_forest • iterations: 100%|████████████████████████████████████████████████████████| 5/5 [01:16<00:00, 15.34s/it]\n", "decision_tree • iterations: 100%|████████████████████████████████████████████████████████| 5/5 [00:39<00:00, 7.96s/it]\n", "xgboost • iterations: 100%|██████████████████████████████████████████████████████████████| 5/5 [01:44<00:00, 20.85s/it]\n", "knn • iterations: 100%|██████████████████████████████████████████████████████████████████| 5/5 [04:04<00:00, 48.93s/it]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "runtime: 1307.1047189235687\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "start_time = time.time() # record running time\n", "results = my_nestedcv_multi(\n", " predictors, y,\n", " n_iter=n_iter, # e.g., repeat the outer loop 3 times\n", " n_jobs=n_jobs,\n", " permutation=False,\n", " threshold_mode=\"youden\",\n", " pca_predictors=False,\n", " # model_names=(\"lasso\",\"ridge\",\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\"),\n", " model_names=(\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\"),\n", " use_gpu=False,\n", " ratio=1.0\n", ")\n", "\n", "\n", "results_permu = my_nestedcv_multi(\n", " predictors, y,\n", " n_iter=n_iter, # e.g., repeat the outer loop 3 times\n", " n_jobs=n_jobs,\n", " permutation=True,\n", " threshold_mode=\"youden\",\n", " pca_predictors=False,\n", " # model_names=(\"lasso\",\"ridge\",\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\"),\n", " model_names=(\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\"),\n", " use_gpu=False,\n", " ratio=1.0\n", ") # permutation\n", "\n", "\n", "end_time = time.time()\n", "runtime = end_time - start_time\n", "print(f\"runtime: {runtime}\")" ] }, { "cell_type": "code", "execution_count": 12, "id": "41abb7dc-c67c-45cb-baca-a0f73519623d", "metadata": {}, "outputs": [], "source": [ "# save results\n", "import pickle\n", "\n", "# save\n", "with open(\"ML_results.pkl\", \"wb\") as f:\n", " pickle.dump(results, f)\n", "\n", "with open(\"ML_results_permu.pkl\", \"wb\") as f:\n", "pickle.dump(results_permu, f)\n", "\n", "# # load back\n", "# with open(\"results.pkl\", \"rb\") as f:\n", "# loaded = pickle.load(f)\n", "# print(loaded)" ] }, { "cell_type": "markdown", "id": "c74b5d52-055b-4b45-ab75-6b6e94274962", "metadata": {}, "source": [ "## Display the model performance" ] }, { "cell_type": "raw", "id": "e77f1826-3485-4a12-ae8e-8b5560a5159b", "metadata": {}, "source": [ "Utilities for summarizing model performance with permutation-based p-values and\n", "compact display strings.\n", "\n", "Key pieces:\n", "- `mypermutation_ttest`: simple permutation comparison of performance vs. chance\n", "- `_mean_ci_95`: mean + 95% CI \n", "- `_stars_from_p`: significance stars from p-values\n", "- `build_model_performance_table`: assemble per-metric, per-model tables ready for display or downstream analysis " ] }, { "cell_type": "code", "execution_count": 26, "id": "ce8fe49b-edc2-4726-ae8d-8d5fe6de4ce7", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import pandas as pd\n", "import scipy.stats as st\n", "\n", "# ---- your original test (unchanged) ----\n", "def mypermutation_ttest(x, y):\n", " \"\"\"Compute a simple permutation-style p-value vs. chance for metric arrays.\n", " \n", " \n", " Parameters\n", " ----------\n", " x : array-like\n", " Permutation distribution for a metric (e.g., scores from label shuffles).\n", " Interpreted as the **null** distribution.\n", " y : array-like\n", " Observed metric values under the real labels.\n", " \n", " \n", " Returns\n", " -------\n", " float\n", " Proportion of permutation samples greater than the **median** of `y`.\n", " (Matches the original behavior precisely.)\n", "\n", " - This is not a full t-test; consider a standard permutation test comparing means or AUC with a two-sided tail if needed\n", " \"\"\"\n", " permutation_large_sum = sum(x > np.median(y))\n", " permutation_size = len(x)\n", " p_value = permutation_large_sum / permutation_size\n", " return p_value\n", "\n", "# ---- helpers for stars / summary formatting ----\n", "\n", "def _stars_from_p(p):\n", " \"\"\"Map a p-value to significance stars.\n", " \"\"\"\n", " if p is None or np.isnan(p):\n", " return \"\"\n", " if p > 0.05:\n", " return \"ns\"\n", " elif p <= 0.0001:\n", " return \"***\"\n", " elif p <= 0.01:\n", " return \"**\"\n", " else:\n", " return \"*\"\n", "\n", "def _mean_ci_95(arr):\n", " \"\"\"Return (mean, [low, high]) using a normal-approximation 95% CI on the mean.\n", " \n", " \n", " Parameters\n", " ----------\n", " arr : array-like\n", " Input values (NaNs ignored).\n", " \n", " \n", " Returns\n", " -------\n", " tuple\n", " `(mean, [ci_low, ci_high])` where the CI is computed as a **normal**\n", " 95% CI on the mean using the standard error of the mean (SEM). Values\n", " are rounded to 3 decimals.\n", " \"\"\"\n", " arr = np.asarray(arr, dtype=float)\n", " arr = arr[~np.isnan(arr)]\n", " if arr.size == 0:\n", " return (np.nan, [np.nan, np.nan])\n", " if arr.size == 1:\n", " m = float(arr[0])\n", " return (m, [m, m])\n", " m = float(np.mean(arr))\n", " # 95% CI for mean (your previous approach): normal approx\n", " ci_lo, ci_hi = st.norm.interval(confidence=0.95, loc=np.mean(arr), scale=st.sem(arr))\n", " return (m, [float(np.round(ci_lo, 3)), float(np.round(ci_hi, 3))])\n", "\n", "# Map internal metric keys to display names/order\n", "_METRIC_ORDER = [\n", " (\"accuracy\", \"ACC\"),\n", " (\"sensitivity\", \"SENS\"),\n", " (\"specificity\", \"SPEC\"),\n", " (\"AUC\", \"AUC\"),\n", " (\"PPV\", \"PPV\"),\n", " (\"NPV\", \"NPV\"),\n", "]\n", "\n", "def build_model_performance_table(\n", " results_by_model,\n", " permutation_results_by_model=None,\n", " models=None,\n", " fmt=\"string\" # \"string\" -> \"mean★, 95%CI([lo, hi])\"; or \"numbers\" -> separate numeric tables\n", "):\n", " \"\"\"\n", " Build a single table with rows=metrics and columns=models.\n", " - results_by_model: dict like results['lasso'] = {\"performance\": {\"score\": {...}}}\n", " - permutation_results_by_model: same structure; if given, stars are computed vs permutation per metric/model\n", " - models: optional list to control column order; defaults to keys in results_by_model\n", " - fmt: \"string\" for compact strings in cells, or \"numbers\" to also return numeric matrices\n", " \n", " Returns:\n", " table_df (if fmt=\"string\")\n", " OR\n", " {\"display\": DataFrame of strings,\n", " \"mean\": DataFrame of means,\n", " \"cilow\": DataFrame of CI lows,\n", " \"cihigh\": DataFrame of CI highs,\n", " \"p\": DataFrame of p-values,\n", " \"stars\": DataFrame of stars}\n", " \"\"\"\n", " if models is None:\n", " models = [m for m in results_by_model.keys()]\n", " \n", " # Prepare containers\n", " display = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=object)\n", " mean_df = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=float)\n", " lo_df = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=float)\n", " hi_df = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=float)\n", " p_df = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=float)\n", " star_df = pd.DataFrame(index=[lab for _, lab in _METRIC_ORDER], columns=models, dtype=object)\n", "\n", " for model in models:\n", " if model not in results_by_model:\n", " continue\n", " scores = results_by_model[model][\"performance\"][\"score\"]\n", " perm_scores = None\n", " if permutation_results_by_model is not None and model in permutation_results_by_model:\n", " perm_scores = permutation_results_by_model[model][\"performance\"][\"score\"]\n", "\n", " for key, label in _METRIC_ORDER:\n", " data = scores.get(key, [])\n", " mean, ci = _mean_ci_95(data)\n", "\n", " # p-value / stars vs permutation (if provided)\n", " pval, stars = (np.nan, \"\")\n", " if perm_scores is not None and key in perm_scores:\n", " try:\n", " pval = mypermutation_ttest(np.asarray(perm_scores[key]), np.asarray(data))\n", " except Exception:\n", " pval = np.nan\n", " stars = _stars_from_p(pval)\n", "\n", " # fill numeric frames\n", " mean_df.loc[label, model] = round(mean, 3) if pd.notna(mean) else np.nan\n", " lo_df.loc[label, model] = ci[0]\n", " hi_df.loc[label, model] = ci[1]\n", " p_df.loc[label, model] = pval\n", " star_df.loc[label, model] = stars\n", "\n", " # build display string\n", " if pd.isna(mean):\n", " disp = \"NA\"\n", " else:\n", " disp = f\"{round(mean,3)}{stars}, 95%CI([{ci[0]}, {ci[1]}])\"\n", " display.loc[label, model] = disp\n", "\n", " if fmt == \"numbers\":\n", " return {\n", " \"display\": display,\n", " \"mean\": mean_df,\n", " \"cilow\": lo_df,\n", " \"cihigh\": hi_df,\n", " \"p\": p_df,\n", " \"stars\": star_df,\n", " }\n", " else:\n", " return display\n" ] }, { "cell_type": "code", "execution_count": 27, "id": "af3c284f-bed2-4b44-881d-2b58f0783fb4", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "predictors → outcomes\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
svmlogisticrandom_forestdecision_treexgboostknn
ACC0.66ns, 95%CI([0.637, 0.684])0.553ns, 95%CI([0.517, 0.589])0.768***, 95%CI([0.726, 0.81])0.754***, 95%CI([0.73, 0.779])0.729***, 95%CI([0.716, 0.742])0.654***, 95%CI([0.614, 0.695])
SENS0.536ns, 95%CI([0.418, 0.654])0.592ns, 95%CI([0.472, 0.712])0.744***, 95%CI([0.656, 0.832])0.544ns, 95%CI([0.513, 0.575])0.712***, 95%CI([0.615, 0.809])0.656ns, 95%CI([0.568, 0.744])
SPEC0.709ns, 95%CI([0.655, 0.764])0.538ns, 95%CI([0.467, 0.609])0.777***, 95%CI([0.731, 0.823])0.838***, 95%CI([0.797, 0.878])0.737***, 95%CI([0.694, 0.78])0.654***, 95%CI([0.593, 0.716])
AUC0.685ns, 95%CI([0.652, 0.719])0.628ns, 95%CI([0.602, 0.655])0.856***, 95%CI([0.82, 0.892])0.697***, 95%CI([0.678, 0.716])0.811***, 95%CI([0.784, 0.837])0.73***, 95%CI([0.683, 0.778])
PPV0.488***, 95%CI([0.431, 0.544])0.349ns, 95%CI([0.313, 0.385])0.6***, 95%CI([0.539, 0.662])0.568***, 95%CI([0.472, 0.664])0.557ns, 95%CI([0.516, 0.598])0.515***, 95%CI([0.439, 0.59])
NPV0.819ns, 95%CI([0.784, 0.855])0.783ns, 95%CI([0.738, 0.828])0.895***, 95%CI([0.855, 0.935])0.834***, 95%CI([0.819, 0.849])0.885***, 95%CI([0.855, 0.914])0.85***, 95%CI([0.814, 0.886])
\n", "
" ], "text/plain": [ " svm logistic \\\n", "ACC 0.66ns, 95%CI([0.637, 0.684]) 0.553ns, 95%CI([0.517, 0.589]) \n", "SENS 0.536ns, 95%CI([0.418, 0.654]) 0.592ns, 95%CI([0.472, 0.712]) \n", "SPEC 0.709ns, 95%CI([0.655, 0.764]) 0.538ns, 95%CI([0.467, 0.609]) \n", "AUC 0.685ns, 95%CI([0.652, 0.719]) 0.628ns, 95%CI([0.602, 0.655]) \n", "PPV 0.488***, 95%CI([0.431, 0.544]) 0.349ns, 95%CI([0.313, 0.385]) \n", "NPV 0.819ns, 95%CI([0.784, 0.855]) 0.783ns, 95%CI([0.738, 0.828]) \n", "\n", " random_forest decision_tree \\\n", "ACC 0.768***, 95%CI([0.726, 0.81]) 0.754***, 95%CI([0.73, 0.779]) \n", "SENS 0.744***, 95%CI([0.656, 0.832]) 0.544ns, 95%CI([0.513, 0.575]) \n", "SPEC 0.777***, 95%CI([0.731, 0.823]) 0.838***, 95%CI([0.797, 0.878]) \n", "AUC 0.856***, 95%CI([0.82, 0.892]) 0.697***, 95%CI([0.678, 0.716]) \n", "PPV 0.6***, 95%CI([0.539, 0.662]) 0.568***, 95%CI([0.472, 0.664]) \n", "NPV 0.895***, 95%CI([0.855, 0.935]) 0.834***, 95%CI([0.819, 0.849]) \n", "\n", " xgboost knn \n", "ACC 0.729***, 95%CI([0.716, 0.742]) 0.654***, 95%CI([0.614, 0.695]) \n", "SENS 0.712***, 95%CI([0.615, 0.809]) 0.656ns, 95%CI([0.568, 0.744]) \n", "SPEC 0.737***, 95%CI([0.694, 0.78]) 0.654***, 95%CI([0.593, 0.716]) \n", "AUC 0.811***, 95%CI([0.784, 0.837]) 0.73***, 95%CI([0.683, 0.778]) \n", "PPV 0.557ns, 95%CI([0.516, 0.598]) 0.515***, 95%CI([0.439, 0.59]) \n", "NPV 0.885***, 95%CI([0.855, 0.914]) 0.85***, 95%CI([0.814, 0.886]) " ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" } ], "source": [ "models_in_order = (\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\")\n", "\n", "\n", "# Do the same for EEG-only and Pedi-only\n", "table_predictors = build_model_performance_table(results, results_permu , models_in_order, fmt=\"string\") # perm_eeg\n", "# table_pedi = build_model_performance_table(results_pedi, perm_pedi, models_in_order, fmt=\"string\")\n", "\n", "print(\"\\npredictors → outcomes\")\n", "table_predictors\n", "\n", "\n", "\n" ] }, { "cell_type": "markdown", "id": "beac5a91-09f0-406a-96b4-e554531df243", "metadata": {}, "source": [ "## Plot the SHAP" ] }, { "cell_type": "code", "execution_count": 48, "id": "8ba701ef-44cd-4bd7-ae39-bb9cfbc3505b", "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Featuremean_abs_shap
0Feature_490.664357
1Feature_590.657823
2Feature_470.524258
3Feature_260.404132
4Feature_460.300160
5Feature_410.288114
6Feature_330.228888
7Feature_570.200986
8Feature_350.194673
9Feature_30.182435
10Feature_310.157442
11Feature_270.140048
12Feature_500.132195
13Feature_200.099229
14Feature_160.097026
15Feature_370.082143
16Feature_510.081249
17Feature_210.077274
18Feature_290.075925
19Feature_180.075559
20Feature_390.071748
21Feature_440.069582
22Feature_420.068766
23Feature_170.061193
24Feature_220.059763
25Feature_360.059216
26Feature_20.055967
27Feature_540.054189
28Feature_480.052283
29Feature_450.049294
30Feature_320.047828
31Feature_130.043113
32Feature_40.037824
33Feature_550.036640
34Feature_250.033612
35Feature_10.031376
36Feature_240.026758
37Feature_150.026149
38Feature_400.021900
39Feature_80.020529
40Feature_300.019574
41Feature_60.016959
42Feature_430.014200
43Feature_110.013500
44Feature_50.013499
45Feature_580.010371
46Feature_100.010156
47Feature_340.009229
48Feature_230.007678
49Feature_140.007328
50Feature_280.006232
51Feature_530.005224
52Feature_380.004656
53Feature_190.003142
54Feature_520.002915
55Feature_560.001687
56Feature_90.000141
57Feature_120.000000
58Feature_70.000000
\n", "
" ], "text/plain": [ " Feature mean_abs_shap\n", "0 Feature_49 0.664357\n", "1 Feature_59 0.657823\n", "2 Feature_47 0.524258\n", "3 Feature_26 0.404132\n", "4 Feature_46 0.300160\n", "5 Feature_41 0.288114\n", "6 Feature_33 0.228888\n", "7 Feature_57 0.200986\n", "8 Feature_35 0.194673\n", "9 Feature_3 0.182435\n", "10 Feature_31 0.157442\n", "11 Feature_27 0.140048\n", "12 Feature_50 0.132195\n", "13 Feature_20 0.099229\n", "14 Feature_16 0.097026\n", "15 Feature_37 0.082143\n", "16 Feature_51 0.081249\n", "17 Feature_21 0.077274\n", "18 Feature_29 0.075925\n", "19 Feature_18 0.075559\n", "20 Feature_39 0.071748\n", "21 Feature_44 0.069582\n", "22 Feature_42 0.068766\n", "23 Feature_17 0.061193\n", "24 Feature_22 0.059763\n", "25 Feature_36 0.059216\n", "26 Feature_2 0.055967\n", "27 Feature_54 0.054189\n", "28 Feature_48 0.052283\n", "29 Feature_45 0.049294\n", "30 Feature_32 0.047828\n", "31 Feature_13 0.043113\n", "32 Feature_4 0.037824\n", "33 Feature_55 0.036640\n", "34 Feature_25 0.033612\n", "35 Feature_1 0.031376\n", "36 Feature_24 0.026758\n", "37 Feature_15 0.026149\n", "38 Feature_40 0.021900\n", "39 Feature_8 0.020529\n", "40 Feature_30 0.019574\n", "41 Feature_6 0.016959\n", "42 Feature_43 0.014200\n", "43 Feature_11 0.013500\n", "44 Feature_5 0.013499\n", "45 Feature_58 0.010371\n", "46 Feature_10 0.010156\n", "47 Feature_34 0.009229\n", "48 Feature_23 0.007678\n", "49 Feature_14 0.007328\n", "50 Feature_28 0.006232\n", "51 Feature_53 0.005224\n", "52 Feature_38 0.004656\n", "53 Feature_19 0.003142\n", "54 Feature_52 0.002915\n", "55 Feature_56 0.001687\n", "56 Feature_9 0.000141\n", "57 Feature_12 0.000000\n", "58 Feature_7 0.000000" ] }, "execution_count": 48, "metadata": {}, "output_type": "execute_result" } ], "source": [ "results[\"xgboost\"][\"shap_summary\"]" ] }, { "cell_type": "code", "execution_count": 33, "id": "f566c27a-f0d3-428b-b1d3-e58c76324692", "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Figure saved to SHAP Plot.png\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "import matplotlib.ticker as mtick\n", "\n", "def plot_shap_summary_bars(results_by_model, models=None, topk=10, normalize='sum', save=True, save_filename=None):\n", " \"\"\"\n", " Plot top-k features per model with a consistent 0–1 scale.\n", "\n", " normalize:\n", " - 'sum' -> value = mean_abs_shap / sum(mean_abs_shap) (shares add to 1)\n", " - 'max' -> value = mean_abs_shap / max(mean_abs_shap) (largest bar = 1)\n", " - None -> raw mean_abs_shap (original behavior)\n", " \"\"\"\n", " if models is None:\n", " models = list(results_by_model.keys())\n", "\n", " n = len(models)\n", " fig, axes = plt.subplots(nrows=n, ncols=1, figsize=(8, 3*n), constrained_layout=True)\n", " if n == 1:\n", " axes = [axes]\n", "\n", " for ax, name in zip(axes, models):\n", " if name not in results_by_model:\n", " ax.axis('off')\n", " continue\n", "\n", " df_all = results_by_model[name][\"shap_summary\"].copy()\n", "\n", " # compute normalization using ALL features, not just top-k\n", " if normalize == 'sum':\n", " denom = df_all[\"mean_abs_shap\"].sum()\n", " # avoid division by zero\n", " df_all[\"norm_value\"] = df_all[\"mean_abs_shap\"] / denom if denom > 0 else 0.0\n", " xlabel = \"share of total |SHAP|\"\n", " xlim = (0, 1)\n", " formatter = mtick.PercentFormatter(1.0) # show as %\n", " elif normalize == 'max':\n", " denom = df_all[\"mean_abs_shap\"].max()\n", " df_all[\"norm_value\"] = df_all[\"mean_abs_shap\"] / denom if denom > 0 else 0.0\n", " xlabel = \"normalized |SHAP| (max=1)\"\n", " xlim = (0, 1)\n", " formatter = None\n", " else:\n", " df_all[\"norm_value\"] = df_all[\"mean_abs_shap\"]\n", " xlabel = \"mean |SHAP|\"\n", " xlim = None\n", " formatter = None\n", "\n", " # pick top-k for plotting\n", " ss = df_all.sort_values(\"norm_value\", ascending=False).head(topk)\n", "\n", " ax.barh(ss[\"Feature\"][::-1], ss[\"norm_value\"][::-1])\n", " ax.set_title(f\"{name} — top {topk} features\")\n", " ax.set_xlabel(xlabel)\n", " ax.set_ylabel(\"feature\")\n", "\n", " if xlim is not None:\n", " ax.set_xlim(*xlim)\n", " if formatter is not None:\n", " ax.xaxis.set_major_formatter(formatter)\n", " \n", " if save and save_filename is None:\n", " save_filename = f\"SHAP Plot.png\"\n", " plt.savefig(save_filename, dpi=300, bbox_inches='tight')\n", " print(f\"Figure saved to {save_filename}\") \n", "\n", " plt.show()\n", "\n", "# Example:\n", "plot_shap_summary_bars(\n", " results,\n", " models=(\"svm\",\"logistic\",\"random_forest\",\"decision_tree\",\"xgboost\",\"knn\"),\n", " topk=20,\n", " normalize=None # or 'max' or None\n", ")\n", "\n" ] }, { "cell_type": "markdown", "id": "9f9af49f-439e-4850-8361-88762b4d0285", "metadata": {}, "source": [ "## Plot the ROC curve" ] }, { "cell_type": "code", "execution_count": 53, "id": "fc0d66f0-7a0a-40be-9e7a-53fe4dd341fd", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "\n", "def _interp_on_grid(fpr, tpr, grid):\n", " \"\"\"Interpolate a single ROC curve (FPR, TPR) onto a shared FPR grid.\n", "\n", " \n", " Parameters\n", " ----------\n", " fpr : array-like\n", " False Positive Rate values from an ROC curve, assumed sorted ascending.\n", " tpr : array-like\n", " True Positive Rate values corresponding to each FPR.\n", " grid : array-like\n", " Common grid of FPR points (e.g., from 0 to 1) where TPR should be interpolated.\n", " \n", " \n", " Returns\n", " -------\n", " np.ndarray\n", " Interpolated TPR values evaluated at the given grid.\n", " \n", " \n", " Notes\n", " -----\n", " Ensures the curve starts at (0,0) and ends at (1,1) before performing\n", " piecewise-linear interpolation using ``np.interp``.\n", " \"\"\"\n", " fpr = np.asarray(fpr, float)\n", " tpr = np.asarray(tpr, float)\n", " if fpr[0] > 0.0:\n", " fpr = np.r_[0.0, fpr]; tpr = np.r_[0.0, tpr]\n", " if fpr[-1] < 1.0:\n", " fpr = np.r_[fpr, 1.0]; tpr = np.r_[tpr, 1.0]\n", " return np.interp(grid, fpr, tpr)\n", "\n", "def _adaptive_fpr_grid(fpr_list):\n", " \"\"\"Generate an adaptive FPR grid based on all ROC curves.\n", " \n", " \n", " Parameters\n", " ----------\n", " fpr_list : list of array-like\n", " List of FPR arrays (one per ROC curve / iteration).\n", " \n", " \n", " Returns\n", " -------\n", " np.ndarray\n", " Sorted, unique array containing the union of all FPR points across iterations,\n", " including 0.0 and 1.0.\n", " \n", " \n", " Notes\n", " -----\n", " This avoids choosing an arbitrary ``grid_size`` and ensures the mean/SEM\n", " calculations preserve all original ROC anchor points.\n", " \"\"\"\n", " all_fpr = np.concatenate([\n", " np.asarray(fpr, float).ravel() for fpr in fpr_list if len(fpr) > 0\n", " ])\n", " all_fpr = np.r_[all_fpr, 0.0, 1.0]\n", " grid = np.unique(np.clip(all_fpr, 0.0, 1.0))\n", " return grid\n", "\n", "def plot_mean_roc_with_sem_adaptive(results_by_model, models=None, ax=None,\n", " label_fmt=\"{model} (AUC≈{auc:.3f})\",\n", " save=True, save_filename=\"ROC_mean_sem.png\"):\n", " \"\"\"Plot mean ROC curves with SEM bands using an adaptive FPR grid.\n", " \n", " \n", " Parameters\n", " ----------\n", " results_by_model : dict\n", " Output dictionary from ``my_nestedcv_multi`` containing ROC data for each model.\n", " models : list[str] or None, optional\n", " Which model keys to plot. Defaults to all keys in ``results_by_model``.\n", " ax : matplotlib.axes.Axes or None, optional\n", " Axis object to draw on. Creates a new figure if None.\n", " label_fmt : str, optional\n", " Format string for legend labels (receives model and AUC).\n", " save : bool, optional\n", " If True, saves the plot as a PNG file.\n", " save_filename : str, optional\n", " Filename for saving the figure if ``save=True``.\n", " \n", " \n", " Returns\n", " -------\n", " matplotlib.axes.Axes\n", " The axis containing the plotted ROC curves.\n", " \n", " \n", " Notes\n", " -----\n", " - Computes the mean and SEM (standard error of the mean) across iterations.\n", " - Uses a data-driven adaptive grid formed from the union of all FPRs.\n", " - Avoids the need for a fixed grid size (e.g., 201 points).\n", " - Draws a shaded SEM band around the mean ROC curve for uncertainty visualization.\n", " \"\"\"\n", " if models is None:\n", " models = list(results_by_model.keys())\n", " if ax is None:\n", " fig, ax = plt.subplots()\n", "\n", " for model in models:\n", " perf = results_by_model[model][\"performance\"]\n", " fpr_list = perf[\"fpr\"] # list of arrays (one per iteration)\n", " tpr_list = perf[\"tpr\"]\n", "\n", " # Build adaptive grid from all iterations’ FPRs\n", " fpr_grid = _adaptive_fpr_grid(fpr_list)\n", "\n", " # Interpolate each iteration onto the adaptive grid\n", " tpr_interp = [\n", " _interp_on_grid(fpr_i, tpr_i, fpr_grid)\n", " for fpr_i, tpr_i in zip(fpr_list, tpr_list)\n", " ]\n", " tpr_interp = np.column_stack(tpr_interp) # (len(fpr_grid), n_iter)\n", "\n", " # Mean and SEM across iterations\n", " mean_tpr = np.mean(tpr_interp, axis=1)\n", " if tpr_interp.shape[1] > 1:\n", " std_tpr = np.std(tpr_interp, axis=1, ddof=1)\n", " sem_tpr = std_tpr / np.sqrt(tpr_interp.shape[1])\n", " else:\n", " sem_tpr = np.zeros_like(mean_tpr)\n", "\n", " # AUC of the mean curve\n", " # auc_mean = np.trapz(mean_tpr, fpr_grid) # calculate the auc from the interpolate values\n", " auc_values = results_by_model[model][\"performance\"][\"score\"][\"AUC\"] # get the auc from the model validation resutls\n", " auc_mean = np.mean(auc_values)\n", "\n", " # Plot mean + SEM band\n", " ax.plot(fpr_grid, mean_tpr, linewidth=2, label=label_fmt.format(model=model, auc=auc_mean))\n", " ax.fill_between(fpr_grid,\n", " np.clip(mean_tpr - sem_tpr, 0, 1),\n", " np.clip(mean_tpr + sem_tpr, 0, 1),\n", " alpha=0.2, linewidth=0)\n", "\n", " ax.plot([0,1],[0,1],'--',linewidth=1,label='Chance')\n", " ax.set_xlabel(\"False Positive Rate\")\n", " ax.set_ylabel(\"True Positive Rate\")\n", " ax.set_title(\"ROC (mean across iterations) with SEM bands — adaptive grid\")\n", " ax.legend(loc=\"lower right\")\n", " ax.set_xlim(0, 1); ax.set_ylim(0, 1)\n", "\n", " if save:\n", " plt.savefig(save_filename, dpi=300, bbox_inches='tight')\n", " print(f\"Figure saved to {save_filename}\")\n", "\n", " plt.show()\n", " return ax\n" ] }, { "cell_type": "code", "execution_count": 55, "id": "b5847c8a-130f-4972-87ff-121e6f82fb98", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Figure saved to ROC_mean_sem.png\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "" ] }, "execution_count": 55, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Plot all models Or select a subset / customize\n", "plot_mean_roc_with_sem_adaptive(results, models=[\"svm\",\"random_forest\",\"decision_tree\",\"xgboost\"])" ] } ], "metadata": { "kernelspec": { "display_name": "python10", "language": "python", "name": "python10" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.18" } }, "nbformat": 4, "nbformat_minor": 5 }