quickOutlier 0.1.5

• Added a NEWS.md file to track package changes.

Functions

• detect_categorical_outliers(): Detects low-frequency outliers in categorical variables based on a percentage threshold.
• detect_lof(): Implements density-based outlier detection using the Local Outlier Factor (LOF) algorithm (via dbscan).
• detect_iforest(): Detects outliers using the Isolation Forest algorithm (via isotree), effective for high-dimensional data.
• detect_multivariate(): Identifies multivariate outliers using Mahalanobis distance with a Chi-square threshold.
• detect_outliers_univ(): Performs univariate outlier detection using either Z-score or Interquartile Range (IQR) methods.
• detect_ts_outliers(): Identifies anomalies in time series data using STL decomposition.
• diagnose_influence(): Diagnoses influential observations in linear regression models using Cook’s distance.
• plot_interactive(): Creates interactive scatter plots using plotly to visualize multivariate outliers.
• plot_outliers(): Generates static ggplot2 visualizations combining boxplots and jittered points to show outliers.
• scan_data(): Scans the entire dataset and provides a summary table of outlier counts and percentages for all numeric columns.
• treat_outliers(): Implements Winsorization (capping) to treat outliers by replacing extreme values with calculated thresholds.

mirror server hosted at Truenetwork, Russian Federation.