Metadata-Version: 1.1
Name: pydistinct
Version: 0.2
Summary: Package for estimating distinct values in a population
Home-page: https://github.com/chanedwin/pydistinct/
Author: Edwin Chan
Author-email: edwinchan@u.nus.edu
License: UNKNOWN
Download-URL: https://github.com/chanedwin/pydistinct/archive/v0.2.tar.gz
Description: # 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
Keywords: distinct,value,estimators,sample,sequences
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
