@eagleoutice/flowr/statistics/summarizer/post-process/histogram.d.ts

Static Dataflow Analyzer and Program Slicer for the R Programming Language

import type { ClusterReport } from './clusterer';
import type { Table } from '../../../util/files';
/**
 * A conventional histogram (e.g., created by {@link histogramFromNumbers}).
 * Can be converted to a {@link Table} by {@link histograms2table}.
 * As described in {@link histogramFromNumbers}, there always will be a special bin for minimum.
 */
export interface Histogram {
    /** A name intended for humans to know what the histogram is about. */
    readonly name: string;
    /** Values located in each bin */
    bins: number[];
    /** The configured size of each bin (stored explicitly to avoid semantic confusion with floating point arithmetic/problems with different rounding schemes) */
    binSize: number;
    /** Minimum value encountered (inclusive minimum of the underlying value range) */
    min: number;
    /** Maximum value encountered (inclusive maximum of the underlying value range) */
    max: number;
    /** Average of the included numbers */
    mean: number;
    /** Standard deviation of the included numbers */
    std: number;
    /** Median of the included numbers */
    median: number;
}
/**
 * Produces column-wise histogram-information based on a {@link ClusterReport}.
 *
 * Let's suppose you want histograms for the Assignments feature.
 * By default, for each clustered value, a histogram is produced (can be configured by `filter`).
 *
 * @param report  - The report to collect histogram information from
 * @param binSize - Size of each bin (see {@link histogramFromNumbers} for details on why we do not specify the bin-count)
 * @param relateValuesToNumberOfLines - If true, each value (like `<-` appeared in file 'x' exactly `N` times) will be divided by the number of lines in the file 'x'.
 * @param filter  - If given, only produce histograms for the given values
 */
export declare function histogramsFromClusters(report: ClusterReport, binSize: number, relateValuesToNumberOfLines: boolean, ...filter: string[]): Histogram[];
/**
 * Produces a histogram from a list of numbers.
 * Because we need to create several histograms of different datasets and want to compare them, we do not accept the
 * number of bins desired and calculate the bin-size from the data (via `Math.ceil((max - min + 1) / bins)`).
 * Instead, we require the bin-size to be given.
 * There *always* will be an extra bin for the minimum value.
 */
export declare function histogramFromNumbers(name: string, binSize: number, values: number[]): Histogram;
/**
 * Takes an array of histograms created by {@link histogramFromNumbers} and produces a CSV table from it.
 * They must have the same bin-size for this function to work.
 *
 * The table has the following columns:
 * - `bin`  - The corresponding bin number
 * - `from` - The exclusive lower bound of the bin
 * - `to`   - The inclusive upper bound of the bin
 * - a column with the name of each histogram, containing its count of values in the corresponding bin
 *
 * @param histograms     - The histogram to convert (assumed to have the same ranges and bins)
 * @param countAsDensity - If true, the count is divided by the total number of values (individually for each histogram, similar to pgfplots `hist/density` option)
 */
export declare function histograms2table(histograms: Histogram[], countAsDensity?: boolean): Table;