essurvey v1.0.8

• Fixes URL creation bug (#53)

essurvey v1.0.7

• CRAN maintenance release. All vignettes are now precompiled to avoid errors when the ESS website breaks for some reason.
• ess_email environmental variable has been renamed to ESS_EMAIL to comply with Github Actions standards

essurvey v1.0.6

essurvey 1.0.6

• CRAN maintenance release to fix Solaris warnings Warning in engine$weave(file, quiet = quiet, encoding = enc) : Pandoc (>= 1.12.3) and/or pandoc-citeproc not available. Falling back to R Markdown v1 on CRAN. Tested on Rhub and all passes OK, notifying CRAN.

• Removes automatic citation message when loading package. It's actually annoying.

essurvey v1.0.5

essurvey 1.0.5

CRAN maintenance release to add more informative message when the status code of the HTTP request of 'www.europeansocialsurvey.org' is more than 300.

Minor changes

All tests/examples are now excluded from running on CRAN based on the warning from Brian Ripley:

'Packages which use Internet resources should fail gracefully with an
informative message if the resource is not available (and not give a check
warning nor error).'

They are all forced to run on Travis and Appveyor and this is made clear on the cran-comments.md

essurvey v1.0.4

Maintenance release for new European Social Survey wave (round 9).

essurvey v1.0.3

essurvey 1.0.3

Breaking changes

• If you don't know which format is available for a round/country, import_* and download_* functions now accept a NULL argument which runs through 'stata', 'spss' and 'sas' formats automatically. By default, import_* functions have now format set to NULL to automatically try the three different formats. This breaks backward dependency but only slightly where it had 'stata' set as default.

New features

• Users can now download SDDF (weight data) for each country/round combination of files. Functions show_sddf_cntrounds, import_sddf_country and download_sddf_country are now introduced. For technical purposes, show_sddf_cntrounds needs for the user to have set their registered ESS email with set_email. [#9]

Minor changes

• Bumps haven to minimum package version 2.1.1
• New package website at https://docs.ropensci.org/essurvey

Internal

• read_format_data now tries to read data using haven but falls backs to foreign in case there's an error. This should only work for SDDF data [#38].
• read_format_data and read_sddf_data now always return a list. Checking the length of data to return a data frame now happens within each import_* function.

Bug fixes

• Removes an unnecessary if statement in set_email that didn't allow to overwrite the email once set.

essurvey v1.0.2

essurvey 1.0.2

Minor changes

show_country_rounds checks if there are missing values and excludes them.

Breaking changes

import_all_cntrounds and import_country returned incorrect countries [#31]

essurvey v1.0.1

MINOR CHANGES

Fixes test that checks the number of rounds that each country has. This test was a mistake because the rounds will change as time passes by and precise country rounds shouldn't be tested.

essurvey v1.0.0

The ess package has been renamed to essurvey for a name conflict with Emacs Speaks Statistics (ESS). See R-pkg mailing list, the post related to the release of ess-0-0-1.

BREAKING CHANGES

• ess_rounds and ess_all_rounds are deprecated and will be removed in the next release. Use import_rounds instead [#22]

• ess_country and ess_all_cntrounds are deprecated and will be removed in the next release. Use import_countries instead [#22]

• The your_email argument name of ess_* functions has be changed to ess_email [#23]

NEW FEATURES

• import_rounds, import_all_rounds and download_rounds have been introduced as
  replacements of ess_rounds and ess_all_rounds. Same changes were repeated for
  ess_country and ess_all_cntrounds [#22]

• set_email to set your email as environmental variable rather than write it in each call [#23]

• All requests to the ESS website are now done through HTTPS rather than HTTP [#24]

• Add package level documentation [#20]

MINOR CHANGES

• ess_email had no default value but now has NULL as default [#23]

• The format argument is now checked through match.arg rathern than manual check [#25]