All functions |
|
---|---|
Generic method to read ADM Datamart data. |
|
Generates PMML files from ADM models. |
|
Turns an ADM Factory JSON into an easy-to-process binning table. |
|
Calculate variable importance from ADM Datamart data. |
|
Sample ADM Data Mart model and predictor data. |
|
Convert AUC performance metric to GINI. |
|
Calculates AUC from counts of positives and negatives directly. |
|
Calculates AUC from an array of truth values and predictions. |
|
Calculates PR AUC (precision-recall) from counts of positives and negatives directly. |
|
Calculates PR AUC (precision-recall) from an array of truth values and predictions. |
|
Default predictor categorization. Return a category label given a predictor name. |
|
Subset the datamart predictor data to only the active predictors. |
|
Subset the datamart predictor data to only the classifier binning. |
|
Subset the datamart predictor data to only the inactive predictors. |
|
Subset the provided datamart data to just the latest snapshot per model. |
|
Remove the binning information from the provided datamart predictor data so it only is predictor level, not more detailed. This is commonly used in plotting functions. |
|
Convert from a Pega date-time string. |
|
Get the active (reachable) score range of one or more models. |
|
Converts a model partition (context) from a JSON representation to a string |
|
Retrieves ADM models from the internal "Factory" table. |
|
Convert a model snapshot into a human-readable as well as machine-executable PMML Scorecard. |
|
Create a human-readable as well as machine-executable Scorecard from the JSON of a specific model. |
|
Check if data has multiple snapshots |
|
A sample export of a Historical Dataset to illustrate how to import and analyze this. |
|
A small sample of IH data to illustrate how to import and analyze this. |
|
Calculates lift from counts of positives and negatives. |
|
pdstools: Open-sourced Data Science tools for Pega. |
|
Create binning plot for any predictor. |
|
Create an alternative binning plot for any predictor. |
|
Create cum gains plot for the model classifier. |
|
Create cum lift plot for the model classifier. |
|
Plot ADM Model Performance over time |
|
Create box plot of performance x success rate from ADM datamart model data. |
|
Create bubble chart from ADM datamart model data. |
|
Create boxplots per predictor with performance range in models. |
|
Performance of predictors per proposition |
|
Plot ADM Proposition Success Rates |
|
Plot ADM Model Success Rate over time |
|
HTML snippet with predictor info. |
|
Read a Pega dataset export file. |
|
PMML Scorecard Reason code specification |
|
PMML Scorecard Reason code specification |
|
PMML Scorecard Reason code specification |
|
PMML Scorecard Reason code specification |
|
Fix field casing |
|
Convert to a Pega date-time string. |
|
Creates user friendly table with binning info. |
|
Write table to a file in the format of the dataset export files. |
|
Calculates the Z-Ratio for the success rate from counts of positives and negatives. |