eventstudyr 1.1.3

This release corresponds to version 1.1.3 on CRAN.

Attached zip file created by devtools::build() from version in commit fd4fccb

What's changed

• Fixed bug when handling unbalanced panels
• Changed CRAN maintainer of package

Relevant issues

• PR for #46 [#48]: Fixed bug when handling of unbalanced panels. Summary here (@santiagohermo, @yang-yuchuan)
• PR for #47 [#49]: Switch maintainer. Summary here (@santiagohermo, @ew487)

Full diff relative to version 1.1.2 here

eventstudyr 1.1.2

This release corresponds to version 1.1.2 on CRAN.

Attached zip file created by devtools::build() from version in commit b0d5bcd

What's changed

• Increased the tolerance to 1e-4 in the solver for finding the smoothest path in order to fix non-convergence error

Relevant issues

• PR for #41 [#43]: Resolved numerical optimization failure in example cases by loosening the default tolerance on the solver minimand. Summary here (@MosesStewart, @cindyjulianelu, @SimonFreyaldenhoven, @jmshapir)
• PR for #44 [#45]: Released to CRAN. Summary here (@rcalvo12, @ew487)

Full diff relative to version 1.1.1 here

eventstudyr 1.1.1

This release corresponds to version 1.1.1 on CRAN.

Attached zip file created by devtools::build() from version in commit d4cd9a9.

What's changed

• Fixed errors in reference manual
• Adjusted the README by adding a reproducible figure, a logo, and badges for CRAN version and number of downloads
• Dropped unused example datasets
• Improved code to compute first differences and leads and lags of policy variable, making it robust to gaps in the time variable in the panel
• Estimated coefficients are now displayed in order from earliest lead to latest lag
• Code is more explicit when handling static models

Relevant issues

• PR for #36 [#37] - Improves reference manual and README, simplifies example datasets. Summary here (@nateschor)
• PR for #31 [#38] - Revise functions PrepareLeads, PrepareLags, and GetFirstDifferences. Summary here (@santiagohermo)
• PR for #39 [#40] - Add eventstudyr version 1.1.1 to CRAN. Summary here (@santiagohermo, @nateschor)

Full diff relative to version 1.0.2 here

eventstudyr 1.0.2

This release corresponds to the first version available on CRAN.

Attached zip file created by devtools::build() from version in commit e430cf5.

What's changed

• No longer export the functions PrepareLeads and PrepareLags
• Remove requirement for a seed in EventStudyPlot, mention that it should be defined outside of calls to eventstudyr functions
• Update citation and vignette
• Improve documentation of functions
• Minor changes to code

Relevant issues

• PR for #28 [#32] - No longer export functions PrepareLeads and PrepareLags, remove seed, introduce citation, proofreading and documentation updates. Summary here (@nateschor, @santiagohermo, @jmshapir)
• PR for #33 [#35] - Submit to CRAN, address feedback. Summary here (@ew487, @nateschor)

Full diff relative to version 1.0.1 here

eventstudyr 1.0.1

This release fixes an important bug while installing the package with devtools::install_github(). It also implements minor fixes and Github Actions.

Zip file created by devtools::build() from version in commit 6508d84.

Bug fixes

• Take /data/ folder out of git lfs, allowing devtools installation
• Improve code structure of /data-raw/ and /tests/testthat/input/
• Implement Github Actions v2

Relevant issues

• PR for #25 [#29] - Fixes installation bug by changing the files in /data/ to no longer use git lfs and adds /data-raw/ (@nateschor)
• PR for #25 [#26] - Improves code in /tests/testthat/input/ (@santiagohermo)
• PR for #27 [#30] - Implements GitHub Actions (@santiagohermo)

Full diff relative to version 1.0.0 here

eventstudyr 1.0.0

Initial release for eventstudyr, an R package that implements tools for estimating linear panel event study models.

Main features:

• Estimate OLS models following Freyaldenhoven et al. (2021)
• Estimate IV models following Freyaldenhoven et al. (2019)
• Construct event-study plots following the suggestions in Freyaldenhoven et al. (2021), including sup-t bands, testing for key hypotheses, and plotting the least wiggly path through the Wald region.