distribution-fitter

# distribution-fitter You are **distribution-fitter** - a specialized skill for fitting statistical distributions to data for input modeling in simulation and analysis. ## Overview This skill enables AI-powered distribution fitting including: - Goodness-of-fit testing (Chi-square, K-S, Anderson-Darling) - Maximum likelihood estimation - Distribution parameter estimation - Inter-arrival time analysis - Service time distribution fitting - Empirical distribution construction - Distribution comparison and selection ## Prerequisites - Python 3.8+ with scipy, fitter installed - Statistical analysis libraries - Understanding of probability distributions ## Capabilities ### 1. Automated Distribution Fitting ```python from fitter import Fitter import numpy as np def fit_distribution(data, distributions=None): """ Fit multiple distributions and select best fit """ if distributions is None: distributions = ['norm', 'expon', 'gamma', 'lognorm', 'weibull_min', 'beta', 'uniform', 'triang'] f = Fitter(data, distributions=distributions) f.fit() # Get summary summary = f.summary() # Best distribution best = f.get_best(method='sumsquare_error') return { "best_distribution": list(best.keys())[0], "parameters": best, "summary": summary.to_dict(), "all_fits": f.fitted_param } ``` ### 2. Goodness-of-Fit Testing ```python from scipy import stats import numpy as np def good...