NEWS.md

In gnoblet/impactR: Ease IMPACT's data and database officers woRk

impactR 0.9.9.9000

• Added: svy_interact() gains the arrange argument, to arrange by proportion highest value first. (#102)
• Bug fix: make_analysis() now uses svy_count_numeric() as a backend for "prop_multiple" and "prop_multiple_overall". (#101)
• New feature: the svy_*() family now output the total unweighted count by group. (#100)

impactR 0.9.8.9000

• New feature: addition of unweighted counts and stat for all svy_*().
• Added: Checks for emptiness and uniqueness of column 'id_analysis' for make_analysis_from_dap() when bind = TRUE.
• Deprecated: all food security indicators functions are deprecated and will be removed in the next iteration. They have been moved to package humind (https://github.com/gnoblet/humind).

impactR 0.9.7.9000

• Bug fix: discrepancy in fcplp() for Phase 5. (#90)

impactR 0.9.6.9000

• Bug fix: label_select_multiple() now get NAs instead of "NA"s.

impactR 0.9.5.9000

• Bug fix: get_choices() correction of a glue error.

impactR 0.9.4.9000

• Bug fix: fclcp() correction for Phase 5, wrong condition before.

impactR 0.9.3.9000

• Bug fix: svy_interact() used across outside of a dplyr-verb, corrected.

impactR 0.9.2.9000

• Bug fix: get_choices now returns an empty character string or tibble if col is not in survey$name.

impactR 0.9.1.9000

• Bug fix: check_analysis is now less restrictive.

impactR 0.9.0.9000

It might be breaking some things. Some rewrite for analysis and checks. Added documentation for analysis

impactR 0.8.11.9000

• Small bug fix: label_columns().
• New feature: label_columns() and get_dictionary() gets a new parameter name_as_label.

impactR 0.8.10.9000

• New feature: it is now possible to label columns (label_columns()) and to get a dictionary of variables from the Kobo survey sheet (get_dictionary()).
• New functions: experimental check functions check_analysis() and check_analysis_dap()
• Add some documentation to make_analysis()

impactR 0.8.9.9000

• Bug fix: bug fix to label_select_one() and label_select_multiple() when there are logical columns.

impactR 0.8.8.9000

• Bug fix: bug fix to make_log_other() when binding others: make sure all "other" columns are of type character.
• Bug fix: bug fix to check_cleaning_log() for identical other_parent_question that are allowed.

impactR 0.8.7.9000

• Bug fix: bug fix to make_analysis_from_dap() when binding if choices for one analysis are numeric and for another one are character.

impactR 0.8.6.9000

• Bug fix: bug fix to check_cleaning_log() for checking identical other old and new values.

impactR 0.8.5.9000

• New feature: welcome to filter2_equalt() and filter2_nequal().

impactR 0.8.4.9000

• Rewrite: make_log_from_check_list() uses purrr::pmap() instead of purrr::exec().

impactR 0.8.3.9000

• New feature: check_cleaning_log() now checks whether the identification column for each survey id_col exists in the cleaning log. It also checks whether all ids from the cleaning log exists in data. (#69)
• New feature: na_count() adds a column with the counts of NAs row-wise over the character columns.

impactR 0.8.2.9000

• Bug fix: for check_cleaning_log().

impactR 0.8.1.9000

• Update: check_cleaning_log() now throw a warning if there are missing 'other_new_value' but 'other_old_value' contained more than 2 values and if there are remaining strings from the template in 'feedback' and 'new_value'.

impactR 0.8.9000

• Breaking: date_to_day_month() loses the useless locale parameter. It now follows the system env locale.
• New: string_count() adds a column with the counts of the string pattern per row over the character columns.

impactR 0.7.12.9000

New features

• make_analysis() and make_analysis_from_dap() gets a lot fresher look and are totally usable. There's still the need for a check_dap() function.

impactR 0.7.11.9000

New features

• get_choices() now returns a tibble when label = TRUE. Two columns, the name and the label. (#61)
• make_analysis() and make_analysis_from_dap() gets a fresh look and can now be used while still under total development. (#62)

impactR 0.7.10.9000

New features

• Four new foodsec indicators calculation: Food Consumption Score fcs(), reduce Coping Strategy Index rcsi(), Household Hunger Scale hhs(), Livelihood Coping Strategy Index lcsi(). (#59)
• Fews net matrix and scores: FEWS NET Food Consumption Matrix's cell fcm(), Food Consumption phase according to the FCM fcp() and FEWS NET Food Consumption-Livelihood Coping phase fclcp(). (#59)
• A general function that calculates all: foodsec(). (#59)
• Thanks to SaeedR1987/healthyr.

impactR 0.7.9.9000

New features

• Two new helpers to export nice excel workbooks using openxlsx: clean_workbook_list() to export a named list of dataframes and clean_workbook_df() to export one dataframe. (#57)
• Updated README.

impactR 0.7.8.9000

New features

• check_cleaning_log() gains many features (#51)
• clean_all() now internally uses check_cleaning_log() before cleaning (#52)
• modify_from_log() now internally uses readr::type_convert() to ensure type conversion of new values from cleaning log (#55)

impactR 0.7.7.9000

Minor bug fixes

• modify_from_log() gains a other argument, it now only modify values if the id_check is different than the other pattern

impactR 0.7.6.9000

Minor bug fixes

• numeric_cols() now do not throw an error when a numeric variable from param survey is not in .tbl colnames. It knows keep only those that are in both (#47)

impactR 0.7.5.9000

Minor bug fixes

• make_all_logs() now throws a warning when there is no column name that starts with the "other" pattern and skips using make_log_other() (#45)

impactR 0.7.4.9000

• Removed all viz functions, now included in package visualizeR (#43)

impactR 0.7.3

Minor bug fixes

• survey_difftime() got a fix for reading the end parameter

impactR 0.7.2

New features

• Redispatched R files (#35)

Breaking changes

• rm_cols() becomes deselect() (#35)
• rec_na() and rec_values() become recode_na() and recode_values() (#35)
• get_one() becomes get_select_one(); get_multiple() becomes get_select_multiple(); and, get_multiple_and_other_parent()becomes get_select_multiple_and_other_parent() (#35)
• svy_difftime() and svy_duration() become survey_difftime() and survey_duration() to avoid confusion with survey analysis svy_* functions (#35)
• survey_duration() gains a new_colname parameter that allows for deciding the new column name of the survey duration (#35)

impactR 0.7.1

Minor bug fixes

• clean_all() and all internals functions now have tidy eval and take into account the "other" string (#30)

impactR 0.7

New features

• Added a NEWS.md file to track changes to the package.
• label() labels all simple and multiple choices column providing kobo data, survey, and choices; label_all_select_one() does it for simple choice columns only, and label_all_select_multiple() for multiple choice columns (#26)
• label_select_one() and label_select_multiple() are helpers, they label one column (#26)

Breaking changes

• get_choices() now has a label parameter to get labelled choices, that takes precedence over the conc parameter (#28)

Minor bug fixes and improvements

• reach_cols() gains a new parameter unnamed = T that allows and has for default an unnamed output vector (#27)
• reach_cols() had fixed hex codes (#27)

gnoblet/impactR documentation built on March 20, 2023, 2:24 a.m.