Statistical Formulas For Programmers

Abstraction: Practical classical statistics formula reference cheat sheet for developers

Key points:

• Covers corrected standard deviation, standard error of the mean, and confidence intervals using t-distribution with N-1 degrees of freedom
• Two-sample t-test for comparing means; Welch degrees of freedom formula accounts for unequal sample sizes and variances
• Confidence interval for Bernoulli parameter uses Wilson score interval (not naive normal approximation)
• Chi-squared test for categorical independence; Poisson confidence intervals via inverse incomplete gamma function
• Kolmogorov-Smirnov tests for one-sample, two-sample, and k-sample distribution comparisons; k-sample critical values require Bessel functions
• Linear regression slope, standard error, and confidence interval formulas included

Connections: Evan Miller · Confidence Intervals · Hypothesis Testing · Kolmogorov Smirnov Test

Source: http://www.evanmiller.org/statistical-formulas-for-programmers.html