Datasets
Models
Downloads: 1
73 lines (73 with data), 9.7 kB
0÷Y│> Ń @ s÷ d Z d d l m Z m Z m Z m Z d d l m Z d d l Z d d l Z d d l
Z d d l Z
d d l Z d d l m Z d d l m Z e d â Z e d â Z d Z d
Z d g Z d Z d Z d Z d
Z d Z Gd d ä d â Z d S)z┼ It is the parent class of the developed training models.
It includes abstract methods, the prediction methods and a set of functions for producing statistical summaries for
the prediction models.
Ú )┌Dict┌List┌TypeVar┌Any)┌metricsN)┌ OrderedDict)┌ CONSTANTS┌ DataFramer zMohsen Mesgarpourz-Copyright 2016, https://github.com/mesgarpour┌GPLz1.1zmohsen.mesgarpour@gmail.com┌Releasec
@ s< e Z d Z d d ä Z e j e e e e e d d d É ä â Z e j e e
d d d É ä â Z e j e e e e d d
d É ä â Z e e e d d
d É ä Z
e e e d d d É ä Z d e e e e
d d d É ä Z d d e e e e e d d d É ä Z e j d d d â e e e e d d d É ä Z d S)!┌Statsc C s) t j t j â | _ | j j t â d S)z.Initialise the objects and constants.
N)┌logging┌ getLoggerr ┌app_name┌_logger┌debug┌__name__)┌selfę r ˙CC:\Users\eagle\Documents\GitHub\Analytics_UoW\TCARER\Stats\Stats.py┌__init__/ s zStats.__init__)┌features_indep_df┌feature_target┌ model_labals┌kwargs┌returnc K s d S)aâ Perform the training, using the defined model.
:param features_indep_df: the independent features, which are inputted into the model.
:param feature_target: the target feature, which is being estimated.
:param model_labals: the target labels (default [0, 1]).
:param kwargs: the training model input arguments.
:return: the trained model.
Nr )r r r r r r r r ┌train5 s
z Stats.train)┌ model_trainr c C s d S)zîProduce the training summary.
:param model_train: the instance of the trained model.
:return: the training summary.
Nr )r r r r r ┌train_summariesD s zStats.train_summaries)r ┌
feature_names┌ class_namesr c C s d S)zˇPlot the tree diagram.
:param model_train: the instance of the trained model.
:param feature_names: the names of input features.
:param class_names: the predicted class labels.
:return: the model graph.
Nr )r r r r r r r ┌plotM s z
Stats.plot)r r r c C s\ | j j d â t â } | j } | j | â | d <| j | â d d ů d f | d <| S)a' Predict probability of labels, using the training model.
:param model_train: the instance of the trained model.
:param features_indep_df: the independent features, which are inputted into the model.
:return: the predicted probabilities, and the predicted labels.
zPredict.┌predNÚ ┌score)r r ┌dict┌values┌predict┌
predict_proba)r r r ┌
model_predictZfeatures_indepr r r r' Z s #z
Stats.predict)r) r r c C s/ | j j d â t â } t j d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d
<t j d | d | d â | d <t j d | d | d â | d <t j d | d | d d
d â | d <t j
d | d | d d
d d d â | d <t j d | d | d â | d <t j d | d | d â | d <t j
d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d <t j d | d | d â | d <| S) a Produce summary statistics for the prediction performance.
:param model_predict: the predicted probabilities, and the predicted labels.
:param feature_target: the target feature, which is being estimated.
:return: the prediction summaries.
z'Produce a prediction summary statistic.┌y_true┌y_predr" ┌accuracy_score┌y_scorer$ ┌average_precision_score┌y_prob┌brier_score_loss┌classification_report┌confusion_matrix┌average┌binary┌f1_score┌betag Ó?┌ fbeta_score┌ hamming_loss┌jaccard_similarity_score┌log_loss┌matthews_corrcoef┌ precision_recall_fscore_support┌precision_score┌ recall_score┌
roc_auc_score┌
zero_one_loss)r r r r r, r. r0 r1 r2 r5 r7 r8 r9 r: r; r< r= r> r? r@ )r r) r ┌ summariesr r r ┌predict_summariesj sF # zStats.predict_summariesg Ó?)┌
predict_scorer ┌cutoffr c s | j j d â t â } t j ç f d d ć | Dâ â } t j d | d | â | d <t j d | d | â | d <t j d | d | â | d
<t j d | d | â | d <t j
d | d | â | d <t j d | d | d
d â | d <t j d | d | d
d d d â | d <t j
d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <t j d | d | â | d <| S) a Produce a shortsummary statistics for the prediction performance.
:param model_predict: the predicted probabilities, and the predicted labels.
:param feature_target: the target feature, which is being estimated.
:return: the prediction summaries.
z-Produce a short prediction summary statistic.c s4 g | ]* } | d k r* | ł k r* d n d Ĺ q S)r r# r )┌.0┌i)rD r r ˙
<listcomp>▓ s z1Stats.predict_summaries_short.<locals>.<listcomp>r* r+ r, r- r. r/ r0 r1 r2 r3 r4 r5 r6 g Ó?r7 r8 r9 r: r; r< r= r> r? r@ )r r r ┌np┌arrayr r, r. r0 r1 r2 r5 r7 r8 r9 r: r; r< r= r> r? r@ )r rC r rD rA Z
predict_labelr )rD r ┌predict_summaries_shortž sH " z Stats.predict_summaries_shortgŕ-üÖŚq=)rC r rD ┌epsilonr c C s§ | j j d â t j d d g d d g d d g d d g d d g d d g d d g d
d g d d g d d g d
d g i â } t | â t | â k rŞ | j j t d â t j â t j d | d | i â } | j d d t j
d d â d t j
d d â i â } t | | d | k | d d k @â } t | | d | k | d d k @â } t | | d | k | d d k @â } t | | d | k | d d k @â }
| | d d <| | d d <| | d d <| | d d <|
| d d <| | t | d â | | d d <| | | | | d d <| | |
| | d
d <| | | | | d d <d | d | | |
| | d d <t j d | d d | d â | d
d <| j
d d â } | j d d d d d d d d
d d d
g d d âS) aÄ Produce short summary statistics for a particular cut-off point.
:param predict_score: the predicted probability.
:param feature_target: the target feature, which is being estimated.
:param cutoff: the risk cut-off point.
:param epsilon: the epsilon value that is used to avoid infinity values.
:return: the summary statistics for the cut-off point.
z;Produce a short prediction summary statistic for a cut-off.rD N┌TP┌FP┌TN┌FN┌Accuracy┌ Precision┌Recall┌ SpecificityzF1-scorezAUC ROCz - different array sizes.r$ ┌target┌dtype┌f2┌i1r# r Ú r* r- ┌dropT┌axis)r r ┌pdr ┌len┌errorr ┌sys┌exit┌astype┌Seriesr r? ┌ reset_index┌ reindex_axis) r rC r rD rK rA ┌df┌tp┌fp┌tn┌fnr r r ┌predict_summaries_cutoffsŠ s6 3$
6(((($&( z Stats.predict_summaries_cutoffsr g═╠╠╠╠╠?gÜÖÖÖÖÖę?)rC r ┌cutoffsr c C sD | j j d â t j d g t | â d g t | â d g t | â d g t | â d g t | â d g t | â d g t | â d g t | â d
g t | â d g t | â d g t | â i â } x, | D]$ } | j | j | | | â â } qÍ W| j d
d â } | j d d d d d d d d d
d d g d d âS)aM Produce a summary statistics table for a range of cut-off points.
:param predict_score: the predicted probability.
:param feature_target: the target feature, which is being estimated.
:param cutoffs: a list of risk cut-off points.
:return: the summary statistics table for the cut-off points.
zBProduce a summary statistics table for a range of cut-off points.rD rL rM rN rO rP rQ rR rS zF1-scorezAUC ROCrY TrZ r# ) r r r[ r r\ ┌appendri rb rc )r rC r rj rA rD r r r ┌ predict_summaries_cutoffs_table s
6 )
" z%Stats.predict_summaries_cutoffs_tableN)r ┌
__module__┌ __qualname__r ┌abc┌abstractmethod┌PandasDataFramer r r r r r! r' ┌CollectionsOrderedDictrB ┌floatrJ ri rH ┌arangerl r r r r r . s(
>!?$6r ) ┌__doc__┌typingr r r r ┌sklearnr r^ ro ┌pandasr[ ┌numpyrH r
┌ collectionsr ┌Configs.CONSTANTSr rq rr ┌
__author__┌
__copyright__┌ __credits__┌ __license__┌ __version__┌__maintainer__┌ __email__┌
__status__r r r r r ┌<module> s( "