pydistinct 0.2

Package for estimating distinct values in a population

# Distinct Value Estimators

This package is based on work from Haas et al, 1995.

It provides a python implementation for the statistical estimators in Haas as well as an XGB ensemble estimator to predict the total number [In progress].

Use case : Given a sample of integers with d distinct values, predict the total number of distinct values D within a population

## Installation

pip install pydistinct

## Usage ` from pydistinct.stats_estimators import * uniform = sampling.sample_gaussian(200,1000,500) print(uniform) >>> {"sample":[-252, -238, -109.. 302, 122], 'sample_distinct': 359, 'ground_truth': 552} bootstrap_estimator(uniform["sample"]) >>> 463.7695215710723 horvitz_thompson_estimator(uniform["sample"]) >>> 519.6486453398398 method_of_moments_v3_estimator(uniform["sample"]) >>> 709.4574356684974 `

## Estimators available : * goodmans_estimator : * chao_estimator : * chao_lee_estimator : * jackknife_estimator : * sichel_estimator : * bootstrap_estimator : * method_of_moments_estimator : * shlossers_estimator : * horvitz_thompson_estimator : * method_of_moments_estimator : * method_of_moments_v2_estimator : * method_of_moments_v3_estimator : * smoothed_jackknife_estimator : * hybrid_estimator :

## Additional planned work

• Include automatic techniques to convert strings to integers

## References