R Packages for Data Science

1. Development
   1. Testing and Exceptions
   2. Utilities
2. Collection
3. Storage
4. Data Structure
   1. Data Frames
   2. Matrices
   3. Time Series
5. Databases
6. Data Transformation
7. Data Validation and Cleaning
8. Data Inspection and Summary
9. Visualization
10. Feature Transformation
    1. Missing Data
    2. Optimal Transforms
    3. Factors
    4. Representation Learning
11. Feature Filtering and Dimension Reduction
12. Feature Selection and Importance
13. Preprocessing
14. Models and Statistics
15. Resampling
16. Hyperparameter Optimization
17. Performance
18. Interpretation
19. Report and Deploy

• devtools: tools for package development
• usethis: automated package tasks
• pacman: package installation/loading utilities
• import: import single functions from a package
• packrat: reproducible development with local package installation
• jetpack: lightweight dependency management for projects and local installs
• checkpoint: install packages from a specific time and date
• config: configuration file management

• testthat:

• assertthat:

• butcher: remove unneeded data from a model
• insight: extract model parameters; see also other packages by the same author
• wrapr: operators and environments
• zeallot: destructuring assignment
• fastpipe: efficient pipe operators

• rio: universal import and export
• opendata: Task View
• tabulizer: pdf table scraping

• RData: native dataframe storage
• RDS: native R objects
• fst: a fast read/write format, like feather
• archivist: collections of objects for reproducibility

• base::data.frame: tabular data
• data.table: efficient data frames
• tibble: user-friendly data frames
• disk.frame: for mid-sized data sets too large to fit in RAM

• Matrix:

• base::ts:

• zoo:

• xts: comprehensive library for timeseries built on zoo

• tidyverts: tidy tools for time series

• tsbox: universal time series converter and utilites

• DBI: for database connections
• dbplyr: dplyr db interface
• rquery: query generator

• dplyr: piped data transformations
• dtplyr: `dplyr` interface to `data.table`
• seplyr: evaluation extensions for dplyr Quoted evaluation, evaluation partitioning.
• cdata: generalized pivots using data description
• forcats: categorical transformations
• stringr: string manipulation

• validate: powerful validation framework with error localization, imputation, and reporting
• janitor: common cleaning operations
• dataMaid: common checks and report generation
• OpenRefine: Java app for cleaning and formatting
• rrefine: R interface to OpenRefine
• assertr: similar to validate but more lightweight
• OfficialStatistics: CRAN view

• inspectdf: text summary and plots

• visdat: data visualization

• UpSetR: intersecting set visualization

• corrplot:

• base::head:

• utils::str:

• dplyr::glimpse:

• base::summary: summary statistics

• skimr: text summary

• HMisc::describe: text summary

• finalfit::ff_glimpse: text summary

• DescTools: Abstract and Describe, text summary and plots

• DataExplorer: text summary and plots, reporting

• base:

• lattice:

• vcd: categorical plots in a lattice framework

• ggplot2: see here for a list of extensions

• cowplot: various extensions to ggplot, themes, formatting, gridplot of all classes

• GGally: various extensions to ggplot, model diagnostics, formatting, and more

• ggExtra: marginal plots for ggplot

• ggpmisc: various extensions, drawing on plots, time series with peaks and valleys , stat plots with embedded statistics

• WVPlots: explanatory and comparison plots in ggplot; "prescribed presentations"

• ggfortify: `autoplot` for a large number of packages

• ggRandomForests:

• naniar: (check out the vignettes, too)
• VIM: imputation of NA data and visualization its distribution
• mice: multiple imputation
• MissMech: tests for the nature of missingness
• finalfit: various plots and reports

• acepack: regression transform selection
• homals: optimal scaling transforms
• aspect: optimal scaling transforms
• vtreat: "y-aware" scaling

• factorMerger: factor response clustering

• vip: variable importance plots
• varImp: variable importance for random forests

• vtreat:

• recipes:

• mlrCPO:

• nns: nonlinear nonparametric statistics with partial moments; classification, clustering and regression, correlation and dependence, forecasting

• forecast:

• smooth:

• vtreat: has CV index generating functions
• sampler:

• mlrMBO:

• IRACE:

• mlr:

• dials:

• performance:

• plotmo: residuals, response, partial dependence

• visreg: regression plots

• car: regression plots

• gamlss::wp: wormplots

• DrWhy: comprehensive collection of tools for exploration and interpretation homepage github - a great resource; lots of interesting and useful packages Predictive Models: Explore, Explain, Debug - book by the authors of DrWhy
• DALEX: local and global interpretation of variables tutorial
• iml: local and global interpretation of variables with interaction support tutorial
• LIME: local interpretation of variables tutorial

• finalfit: creates formatted tables of many types
• dashR: R interface to plotly's Dash framework
• shiny: