PDFModeling Results
Holds all the results from the PDF modeling process.
The captured evaluation model permits the translation of results to data frame representations.
Constructors
Properties
The results sorted by the average ranking of the metrics
The top result based on the average ranking of metrics
The top result according to the scoring evaluation model.
The top result according to the average ranking of the metrics specified as the type of random variable.
The top result according to the scoring evaluation model specified as the type of random variable.
Functions
Returns the metric in the form of a data frame. The first column is the distributions that were evaluated. Subsequent columns represent the metric results for each distribution including the overall metric value. The rows are sorted by overall value.
The PDF modeling process estimates the parameters for a set of parameter estimators. Each estimator is unique in the set of estimators that are processed. Each estimator produces scores on its quality of fit for the distribution. The scores are combined into an overall score for the estimator.
A rank of 0, means that the random variable type rvType was not found in the results. The returned rank {1, 2, ...} is dependent on the number of distributions that were fit within the PDF modeling process. A rank of 1, means that the distribution was the recommended distribution (top ranked) based on the scoring model. Thus, lower ranks imply better distribution fit. The returned rank is the first instance of the random variable type.
Returns a data frame with the first column being the distributions by name, a column of ranks for each metric for each distribution. The metric ranking columns are labeled as "${metric.name}_Rank"
The results and their overall rank based on the average of their metric ranks
The results and their overall rank based on the value scoring model.
Returns the scores in the form of a data frame. The first column is the distributions that were evaluated. Subsequent columns represent the scoring model results for each distribution.