Статистический анализ данных и временных рядов

★ 7.3 · content

statistical-analysis is a Claude Code skill that applies statistical methods to data analysis, covering descriptive statistics (mean, median, percentiles from p1 to p99, IQR, coefficient of variation), trend and seasonality analysis, outlier detection via Z-score and IQR methods, basic forecasting, and hypothesis testing. The skill includes ready-to-use Python code for rolling averages, period-over-period growth rates (WoW, MoM, YoY, CAGR), and time-series anomaly detection. A core emphasis is on sound interpretation: when to prefer median over mean, why outliers should be investigated rather than automatically removed, and how to communicate forecast uncertainty with ranges instead of point estimates. It is designed for analysts working with product or business metrics who need to characterize distributions, test for significance, or surface anomalies in operational data.