Catalog

https://orbisapi.com/proxy/pca-variance-explainer-api-448b19

Detect univariate anomalies using Z-score or IQR method with severity classification.

Calculate DPMO, sigma level, yield and improvement roadmap from defect counts or DPMO input.

Estimate Kaplan-Meier survival curves with confidence intervals from event and censoring data.

Calculate Cp, Cpk, Cpm process capability indices with DPMO, sigma level and Six Sigma grade.

Test Granger causality between two time series across multiple lags

Test Granger causality between two time series across multiple lags. Returns F-statistics per lag.

Calculate exponentially weighted moving averages with configurable alpha

Calculate exponentially weighted moving averages with configurable alpha. Includes MACD calculation.

Detect autocorrelation in time series using ACF, Durbin-Watson and significant lag identification.

Analyze SPC control charts with UCL/LCL, out-of-control detection, Nelson rules and Cpk calculation.

Calculate rolling mean, std dev, min, max, median and sum with configurable window size.

Test regression residuals for heteroscedasticity using variance ratio and residual correlation analysis.

Calculate Variance Inflation Factor per feature to detect multicollinearity in regression models.

https://orbisapi.com/proxy/information-gain-scorer-api-016608

Calculate mutual information and normalized MI between two variables with variation of information.

Calculate Shannon/natural/log10 entropy and normalized entropy for any categorical distribution.

Calculate precision-recall curve and AUC-PR with F-beta optimization at configurable beta value.

Calculate ROC curve and AUC from prediction scores and binary labels

Calculate ROC curve and AUC from prediction scores and binary labels. Returns optimal threshold.

Analyze confusion matrix for precision, recall, F1, macro averages and best/worst class identification.

Rank features by Pearson correlation with target or provided importance scores

Rank features by Pearson correlation with target or provided importance scores. Returns relative importance.

Calculate Gini coefficient and Lorenz curve for any value distribution

Calculate Gini coefficient and Lorenz curve for any value distribution. Returns inequality level.

Score class imbalance severity and recommend SMOTE, class weights or undersampling techniques.

Recommend CV strategy (k-fold, LOO, time series split) from dataset size, time series and imbalance flags.

https://orbisapi.com/proxy/data-drift-detector-api-875a53

Score forecast accuracy with MAE, RMSE, MAPE, bias direction and R-squared.