Crate u_analytics 

u-analytics

Statistical process control (SPC), process capability analysis, Weibull reliability, and change-point detection.

This crate provides industrial quality analysis tools that are domain-agnostic — they operate on raw f64 data without knowledge of manufacturing, scheduling, or any specific consumer domain.

Modules

• spc — Control charts (X̄-R, X̄-S, I-MR, P, NP, C, U) with run rules
• capability — Process capability indices (Cp, Cpk, Pp, Ppk, Cpm)
• weibull — Weibull parameter estimation (MLE, MRR) and reliability analysis
• detection — Change-point detection (CUSUM, EWMA, PELT)
• smoothing — Time series smoothing (SES, Holt, Holt-Winters)
• seasonality — Periodogram and dominant-period estimation (AutoPeriod)
• decomposition — Seasonal-trend decomposition by Loess (STL)
• correlation — Correlation analysis (Pearson, Spearman, Kendall, matrices)
• regression — Regression analysis (simple, multiple OLS, VIF, prediction intervals)
• accuracy — Error measures of predictions (MAE, RMSE, CV(RMSE), NMBE)
• distribution — Distribution analysis (ECDF, histogram bins, QQ-plot, KS test)
• testing — Hypothesis testing (t-tests, ANOVA, chi-squared, normality)
• error — Refusals shared by more than one module (error::TimeSeriesError)
• msa — Measurement System Analysis (Gage R&R, ANOVA method)

Design Philosophy

• Domain-agnostic: No manufacturing or process-specific types
• Numerical stability: Leverages u-numflow for stable statistics
• Research-backed: All algorithms reference academic literature

Modules

accuracy

Accuracy of a model’s predictions against observations.

capability

Process capability analysis.

correlation

Correlation analysis.

counter

Increments of a cumulative counter, each classified by what explains it.

decomposition

Seasonal-trend decomposition of a univariate series by Loess (STL).

detection

Change-point and shift detection.

distribution

Distribution analysis.

error

Refusals shared by the functions of more than one module.

msa

Measurement System Analysis (MSA).

point_process

Trend tests and the power-law fit for the event times of a point process.

regression

Regression analysis.

seasonality

Seasonality: the periodogram of a univariate series and the estimation of its dominant period.

smoothing

Time series smoothing and forecasting.

spc

Statistical Process Control (SPC) charts.

testing

Hypothesis testing.

weibull

Weibull parameter estimation and reliability analysis.

Constants

VERSION

This crate’s version, as published (CARGO_PKG_VERSION) — for a consumer that reports what it was built against without reading a lock file.