R Packages Updated in June 2020

The R ecosystem continues to evolve rapidly, with June 2020 seeing numerous package updates across various domains. This month was particularly interesting as it featured updates to core tidyverse packages, machine learning tools, visualization libraries, and specialized statistical packages. This article highlights some of the most significant updates and discusses their implications for R users and the broader data science community.

Tidyverse Ecosystem Updates

Core Tidyverse Packages

June 2020 brought several meaningful updates to the core tidyverse packages, which form the foundation of modern data science workflows in R:

dplyr 1.0.0 - Major update introducing significant changes to data manipulation workflows
tidyr 1.1.0 - Enhanced reshaping capabilities with new pivot functions
ggplot2 3.3.2 - Plotting improvements and bug fixes
stringr 1.4.0 - New string manipulation functions
readr 1.3.1 - Performance improvements for data import

The most significant of these was undoubtedly dplyr 1.0.0, which marked a milestone in the evolution of data manipulation in R. This version introduced several new features that expanded the package's capabilities while maintaining backward compatibility for most existing code.

dplyr 1.0.0: Revolutionary Changes

The release of dplyr 1.0.0 brought several important improvements:

Introduction of across() for powerful column-wise operations replacing summarise_at(), mutate_at(), and similar functions
Enhanced grouping capabilities with group_by() supporting more complex grouping patterns
Improved handling of list-columns with new cur_data() and cur_group() helpers
Better support for custom functions within verbs
Performance improvements for large datasets

The across() function in particular represented a philosophical shift in how operations across multiple columns could be performed, offering more flexibility and consistency in syntax compared to the previous scoped variants.

Machine Learning Updates

caret and Related Packages

Machine learning practitioners saw several important updates in June 2020:

caret 6.0-86 - Bug fixes and performance improvements
mlr 2.19.0 - Enhanced hyperparameter tuning capabilities
tidymodels framework packages - Multiple updates across the ecosystem
h2o 3.30.0.1 - Integration improvements and new algorithms

These updates reflected the growing maturation of machine learning workflows in R, with increasing emphasis on reproducibility, interoperability between methods, and easier transition from model development to production deployment.

tidymodels Ecosystem

The tidymodels framework, designed to replace and improve upon many traditional machine learning workflows, received significant attention in June 2020:

parsnip 0.1.2 - Improved model specification and workflow integration
recipes 0.1.13 - Enhanced preprocessing capabilities
rsample 0.0.8 - New resampling methods
tune 0.1.1 - Better hyperparameter tuning workflows
workflows 0.1.2 - Simplified model coordination

These updates collectively made the tidymodels approach more accessible and powerful, lowering the barrier to entry for users transitioning from caret or other traditional machine learning frameworks.

Data Visualization Updates

Plotting Enhancements

Data visualization tools received several valuable updates in June 2020:

ggplot2 3.3.2 - Improved axis labeling and bug fixes
plotly 4.9.2.1 - Better integration with ggplot2
patchwork 1.0.1 - Enhanced plot composition
lattice 0.20-41 - Maintenance updates and bug fixes

The ggplot2 update addressed several long-standing issues with axis scaling and labeling that had frustrated users, particularly those working with log-transformed data or custom axis scales.

plotly Enhancements

The plotly package update brought several improvements that made interactive visualization more seamless:

Better conversion from ggplot2 graphics
Improved performance for large datasets
New animation capabilities
Enhanced export functionality for static images

These updates bridged further gaps between static and interactive visualization workflows, making it easier for users to create compelling interactive reports without learning entirely new syntax.

Specialized Domain Updates

Spatial Analysis Packages

The spatial analysis ecosystem saw significant updates in June 2020:

sf 0.9-3 - Enhanced spatial operations and performance improvements
sp Data 0.3.0 - Updated administrative boundary data
terra 0.5-2 - New package for raster data handling
mapview 2.8.0 - Enhanced interactive mapping

The sf package updates focused on spatial operations performance, particularly for complex geometries and large spatial datasets, while also adding new spatial predicates and analysis functions.

Time Series Analysis

Time series specialists received several useful updates:

forecast 8.13 - Improved forecasting algorithms and visualization
timetk 2.5.0 - Enhanced time series feature engineering
xts 0.12-0 - Performance improvements
fable 0.2.1 - Updated modeling framework

These updates continued the evolution of time series analysis in R toward more modern, tidyverse-compatible workflows without sacrificing the advanced statistical modeling capabilities that have made R a preferred environment for time series analysts.

Integration and Infrastructure Updates

Database Connectivity

Database operations in R were improved with several package updates:

DBI 1.1.0 - Enhanced database interface capabilities
odbc 1.3.0 - Improved Windows database connectivity
pool 0.1.6 - Better connection management
RMariaDB 1.0.8 - Updated MySQL/MariaDB interface

These updates focused on making database interactions more reliable and performant, particularly for enterprise applications where connection management and query optimization are critical.

Report Generation

The report generation ecosystem received significant attention:

rmarkdown 2.3 - Enhanced document formatting and output options
bookdown 0.20 - Improved book and long document authoring
flexdashboard 0.5.2 - New layout options and interactivity
xaringan 0.16 - Enhanced presentation capabilities

These updates made it easier to create professional, polished documents directly from R, supporting the growing trend toward reproducible research and automated reporting in data science workflows.

Conclusion

June 2020 represented a period of significant advancement in the R ecosystem. The updates reflected several clear trends:

Greater integration and coherence between packages within ecosystems like tidyverse and tidymodels
Continued emphasis on performance improvements for large datasets
Expanded capabilities for reproducible research and reporting
Moving toward more modern, consistent interfaces while maintaining backward compatibility

These updates collectively improved the R experience for data scientists across a wide range of domains. As the ecosystem continues to mature, we see increasing specialization for particular analytical domains alongside greater standardization of interfaces and workflows within domains. This combination of specialized capabilities and standardized interfaces makes R increasingly powerful while simultaneously becoming more accessible to new users.

The evolution we observed in June 2020 continues to influence how data science workflows are designed. The introduction of functions like dplyr's across() and maturation of the tidymodels framework represent lasting contributions to how data analysis is performed in R, demonstrating both the innovation and stability of the ecosystem.

```

Reference Files For R Packages Updated June 2020

Screenshoot