o xi-@sdZddlmZddlZddlmZddlZddlZddl m Z m Z ddl m Z eded   dd d Z      dddZ      dddZ d ddZ d!ddZd"ddZd#ddZdS)$z?Define plots for classification models built with scikit-learn.) simplefilterN) naive_bayes) calculateutils)sharedignore)actioncategoryF Classifierc Cstd| dt|tjst|| td| r't|||tdt|||tdtj |||||dtdt |||tdt|tjs^t |||| td t |||td t |||td d S) agGenerate all sklearn classifier plots supported by W&B. The following plots are generated: feature importances, confusion matrix, summary metrics, class proportions, calibration curve, roc curve, precision-recall curve. Should only be called with a fitted classifier (otherwise an error is thrown). Args: model: (classifier) Takes in a fitted classifier. X_train: (arr) Training set features. y_train: (arr) Training set labels. X_test: (arr) Test set features. y_test: (arr) Test set labels. y_pred: (arr) Test set predictions by the model passed. y_probas: (arr) Test set predicted probabilities by the model passed. labels: (list) Named labels for target variable (y). Makes plots easier to read by replacing target values with corresponding index. For example if `labels=['dog', 'cat', 'owl']` all 0s are replaced by dog, 1s by cat. is_binary: (bool) Is the model passed a binary classifier? Defaults to False model_name: (str) Model name. Defaults to 'Classifier' feature_names: (list) Names for features. Makes plots easier to read by replacing feature indexes with corresponding names. log_learning_curve: (bool) Whether or not to log the learning curve. Defaults to False. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_classifier( model, X_train, X_test, y_train, y_test, y_pred, y_probas, ["cat", "dog"], False, "RandomForest", ["barks", "drools", "plays_fetch", "breed"], ) ``` z Plotting .zLogged feature importances.zLogged learning curve.zLogged confusion matrix.)XyX_testy_testzLogged summary metrics.zLogged class proportions.zLogged calibration curve.zLogged roc curve.zLogged precision-recall curve.N)wandbtermlog isinstancer MultinomialNBfeature_importancesrlearning_curveconfusion_matrixsummary_metricsclass_proportionscalibration_curverocprecision_recall) modelX_trainry_trainry_predy_probaslabels is_binary model_name feature_nameslog_learning_curver']/home/ubuntu/.local/lib/python3.10/site-packages/wandb/integration/sklearn/plot/classifier.py classifiers(>              r)TcCs$tj||||}td|idS)aLog the receiver-operating characteristic curve. Args: y_true: (arr) Test set labels. y_probas: (arr) Test set predicted probabilities. labels: (list) Named labels for target variable (y). Makes plots easier to read by replacing target values with corresponding index. For example if `labels=['dog', 'cat', 'owl']` all 0s are replaced by dog, 1s by cat. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_roc(y_true, y_probas, labels) ``` rN)rplot roc_curvelog)y_truer!r" plot_micro plot_macroclasses_to_plot roc_chartr'r'r(rnsrc Csft|}t|}tj||d}tj||d}|r/|r1t||||||}td|idSdSdS)aLog a confusion matrix to W&B. Confusion matrices depict the pattern of misclassifications by a model. Args: y_true: (arr) Test set labels. y_probas: (arr) Test set predicted probabilities. labels: (list) Named labels for target variable (y). Makes plots easier to read by replacing target values with corresponding index. For example if `labels=['dog', 'cat', 'owl']` all 0s are replaced by dog, 1s by cat. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_confusion_matrix(y_true, y_probas, labels) ``` )r-r rN) npasarrayr test_missing test_typesrrrr,) r-r r" true_labels pred_labels normalize not_missing correct_typesconfusion_matrix_chartr'r'r(rs   rcCs$tj||||}td|idS)a[Log a precision-recall curve to W&B. Precision-recall curves depict the tradeoff between positive predictive value (precision) and true positive rate (recall) as the threshold of a classifier is shifted. Args: y_true: (arr) Test set labels. y_probas: (arr) Test set predicted probabilities. labels: (list) Named labels for target variable (y). Makes plots easier to read by replacing target values with corresponding index. For example if `labels=['dog', 'cat', 'owl']` all 0s are replaced by dog, 1s by cat. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_precision_recall(y_true, y_probas, labels) ``` rN)rr*pr_curver,)r-r!r"r.r0precision_recall_chartr'r'r(rsrFeature Importance2cCsXtj|d}tj|d}t|}|r&|r(|r*t||}td|idSdSdSdS)aLog a plot depicting the relative importance of each feature for a classifier's decisions. Should only be called with a fitted classifier (otherwise an error is thrown). Only works with classifiers that have a feature_importances_ attribute, like trees. Args: model: (clf) Takes in a fitted classifier. feature_names: (list) Names for features. Makes plots easier to read by replacing feature indexes with corresponding names. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_feature_importances(model, ["width", "height", "length"]) ``` )rrN)rr4r5 test_fittedrrrr,)rr%titlemax_num_featuresr9r: model_fittedfeature_importance_chartr'r'r(rs      rcCsbtj||d}tj||d}|r-|r/t|t|}}t|||}td|idSdSdS)aPlot the distribution of target classes in training and test sets. Useful for detecting imbalanced classes. Args: y_train: (arr) Training set labels. y_test: (arr) Test set labels. labels: (list) Named labels for target variable (y). Makes plots easier to read by replacing target values with corresponding index. For example if `labels=['dog', 'cat', 'owl']` all 0s are replaced by dog, 1s by cat. Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_class_proportions(y_train, y_test, ["dog", "cat", "owl"]) ``` )rrrN) rr4r5r2arrayrrrr,)rrr"r9r:class_proportions_chartr'r'r(rsrcCstj|||d}tj|||d}t|}|rH|rJ|rLt|}|jjdks0|dk|dkBs7t ddSt ||||}t d|idSdSdSdS)aLog a plot depicting how well-calibrated the predicted probabilities of a classifier are. Also suggests how to calibrate an uncalibrated classifier. Compares estimated predicted probabilities by a baseline logistic regression model, the model passed as an argument, and by both its isotonic calibration and sigmoid calibrations. The closer the calibration curves are to a diagonal the better. A sine wave like curve represents an overfitted classifier, while a cosine wave like curve represents an underfitted classifier. By training isotonic and sigmoid calibrations of the model and comparing their curves we can figure out whether the model is over or underfitting and if so which calibration (sigmoid or isotonic) might help fix this. For more details, see https://scikit-learn.org/stable/auto_examples/calibration/plot_calibration_curve.html. Should only be called with a fitted classifier (otherwise an error is thrown). Please note this function fits variations of the model on the training set when called. Args: clf: (clf) Takes in a fitted classifier. X: (arr) Training set features. y: (arr) Training set labels. model_name: (str) Model name. Defaults to 'Classifier' Returns: None: To see plots, go to your W&B run page then expand the 'media' tab under 'auto visualizations'. Example: ```python wandb.sklearn.plot_calibration_curve(clf, X, y, "RandomForestClassifier") ``` )clfr rUrrzThis function only supports binary classification at the moment and therefore expects labels to be binary. Skipping calibration curve.Nr)rr4r5r@r2r3dtypecharallrtermwarnrcalibration_curvesr,)rGr rclf_namer9r: is_fittedcalibration_curve_chartr'r'r(rs!     r)Fr NF)NNNTTN)NNNNNF)NNNTN)NNr>r?)NNN)NNNr )__doc__warningsrnumpyr2sklearnrr wandb.plotwandb.integration.sklearnrrr FutureWarningr)rrrrrrr'r'r'r(sD      ]   1 !