Table 1 Operations and Algorithms implemented in COMPSs available in Lemonade
From: Upgrading a high performance computing environment for massive data processing
Categories | Operations and algorithms |
|---|---|
Data | Read and write files, attributes changer, data balancer |
ETL | Add columns, aggregation, clean missing data, difference, distinct (remove duplicate rows), drop columns, filter, intersection, joins (inner, left and right join), replace values, sample, select columns (projection), sort, split, transformation, union. |
Geographic | Read shapefile, Geo within (check if a point is within a region), ST-DBSCAN |
Graph | PageRank |
Metrics | Classification (accuracy, precision/recall and f-measure), regression (MSE, RMSE, MAE, R2) |
ML | Feature assembler, Scalers (min-max, max-abs and standard), String Indexer, PCA, K-Means, DBSCAN, KNN, Naive Bayes, SVM, Logistic regression, Linear regression, Apriori, Load/Save model |
Text | Vectorization by Bag-of-Words and Tf-idf, tokenizer, stop-words remover |