aabderrahimaariche-hue/DataPreparation: DataPreparation v1.0.0

DataPreparation A comprehensive, modular R package for data preparation and feature engineering. This package provides 50+ general-purpose functions organized into 7 logical modules, enabling reuse across different projects and datasets. Features 📦 Module 1: NA Handling Functions for managing missing values: remove_na_columns() - Remove rows with NA in specified columns impute_na_numeric() - Impute NA with median values flag_na_column() - Create binary NA indicator flags summarize_na_rows() - Generate NA statistics summaries 🔢 Module 2: Numeric Features (30+ functions) Comprehensive numeric feature engineering: Scaling: scale_betweenRange(), scale_unbounded() Transformations: transform_log(), transform_sqrt(), transform_exp() Polynomial Features: create_polynomial_features(), create_interaction_features(), create_ratio_features() Binning: binning_quantile(), binning_width() Outlier Detection: flag_outliers_iqr(), flag_outliers_zscore() Normalization: normalize_numeric() (4 methods: minmax, zscore, robust, log) Statistics: compute_rolling_statistics(), compute_group_statistics() Advanced Features: create_quantile_features(), create_rank_features(), create_percentage_change(), create_difference_features() Variants: generate_numeric_variants(), generate_numeric_variants_advanced() ⏰ Module 3: Timestamp Features Extract temporal patterns: extract_timestamp_features() - Extract year/month/day/weekday from timestamps extract_year_from_title() - Extract 4-digit year from string titles 📝 Module 4: String Features String processing and feature validation: validate_ngram() - Validate n-gram extraction (clean vs. distorted) generate_numeric_variants() - Create numeric feature transformation candidates generate_categorical_variants() - Generate categorical encoding variants (one-hot encoding) 🔤 Module 5: N-gram Generation Advanced text feature extraction: generate_ngram_char_variants() - Character-level n-grams using 4 methods: Sliding: Overlapping n-grams (step by 1) Stride: Non-overlapping n-grams excluding spaces Strict: Non-overlapping with underscore padding Null: Non-overlapping, incomplete chunks removed generate_ngram_word_variants() - Word-level n-grams (unigrams, bigrams, trigrams, etc.) 📍 Module 6: Position Extraction Extract position-specific patterns from n-grams: extract_prefix_patterns() - Extract first/opening n-grams (prefix position) extract_infix_patterns() - Extract middle n-grams (infix position) extract_suffix_patterns() - Extract last/closing n-grams (suffix position) 🎯 Module 7: Linguistic Analysis (Dynamic Pattern Discovery) Advanced linguistic feature extraction with dynamic pattern discovery: extract_string_linguistic_profile() - Extract structural and dynamic linguistic features compute_global_linguistic_effects() - Compute feature effects on target variable compute_user_linguistic_preferences() - Extract user-specific feature preferences compute_user_linguistic_compatibility() - Calculate user-feature compatibility scores compute_position_importance() - Rank position importance (prefix, infix, suffix) summarize_linguistic_analysis() - Generate comprehensive analysis summaries