v0.1.4

wbpip 0.1.4

• update webpage

wbpip 0.1.3

main changes

1. Key values: removed as individual arguments, instead supplied as one list using argument key_values created using gd_lq_key_values

2. Remove unnecessary dd argument from gd_compute_mld_* and gd_compute_watts_*

3. This is potentially too much information, but here it is. The function check_curve_validity_lq() is called in DEV in the following ways. It is used a) line 115 in gd_compute_dist_stats.R in gd_estimate_dist_stats_lq(), where the r argument is squared as the input; b) line 784 in gd_compute_pip_stats_lq.R in gd_estimate_lq() where the r argument is not squared; c) in line 137 of gd_compute_poverty_stats.R in function gd_estimate_poverty_stats_lq() where the r argument is not squared; d) line 103 of prod_gd_compute_pip_stats_lq().R in function prod_gd_estimate_lq() where the r argument is not squared. Therefore, out of the four times it is called it is only the first where the r argument is squared. The paper (Villasenor & Arnold, 1989) uses the square root for the r argument, similar (although with a denominator) to r calculated in gd_lq_key_values. However, the other three calls of check_curve_validity_lq() mentioned above do not calculate the square root for r prior to the function call , which is why they just use the r argument directly. What has been done here is to i) use the square root for r in key_values list; ii) but then square the r argument within check_curve_validity_lq(). However, r is never actually used anywhere in check_curve_validity_lq() except in if(r 0) {return(...)} as an early check. This requires a bit more investigation to see whether the r argument is necessary here.

4. {collapse} - use functions directly, and use max() and sum() rather than fmax() and fsum() because base functions are more efficient.

5. quantile functions in md_compute_quantiles.R that are efficient, utilize lorenz argument, and wrap around md_compute_lorenz()

6. Improvements in efficiency are made to gd_compute_mld_lq() and gd_compute_quantile_lq(). However, there are no changes in relation to the bug fixes yet.

7. Removed the functions in md_quantile_functions.R. First ensured that a corresponding function that served the same purpose could be found in md_compute_quantiles.R

8. increase efficiency in md_compute_polarization and correct the tests. The benchmarks in the tests were slightly off (at around the fourth decimal). When using the formula to calculate the benchmarks analytically, the tests pass.

Some additional/general comments:

• Exported many additional functions.
• Some functions have been improved/vectorized. The new functions have taken the names of the old functions, and the old functions have the prefix "old_"

v0.1.2

What's Changed

• Tests and bugs by @randrescastaneda in #224
• improve documentation and small changes by @randrescastaneda in #226
• functions to estimate low-hanging fruit indicators. Those that use standard output as input. by @randrescastaneda in #225
• Qa by @tonyfujs in #228
• Unit tests by @dana89co in #234
• Unit test md by @dana89co in #235
• Refactor_consistency_review_AC by @randrescastaneda in #238
• Md welfare share at alternative by @giorgiacek in #241
• Refactor consistency by @zander-prinsloo in #236
• Merge QA by @tonyfujs in #249

New Contributors

• @giorgiacek made their first contribution in #241

Full Changelog: v0.1.0...v0.1.2

v0.1.0

What's Changed

• vectorize spl by @randrescastaneda in #216
• small optimization by @tonyfujs in #222
• Dev by @tonyfujs in #223

Full Changelog: v0.0.4...v0.1.0

v0.0.4

What's Changed

• Add synthetic microdata by @Aeilert in #6
• Add md_compute_gini() by @Aeilert in #7
• Md check data by @randrescastaneda in #8
• Popshare by @randrescastaneda in #18
• Quantiles by @randrescastaneda in #19
• add md_compute_lorenz() function by @tonyfujs in #15
• Ifynew by @ifeanyi588 in #34
• add md_compute_mld() by @tonyfujs in #17
• Gd lq by @tonyfujs in #24
• Md compute pip stats by @tonyfujs in #38
• Add function to deflate welfare means by @Aeilert in #42
• Update Lorenz Quadratic code (formatting and minor refactoring) by @tonyfujs in #50
• Add fill_gaps() by @Aeilert in #43
• First version of gd_clean_data() by @randrescastaneda in #49
• Make sure data is sorted by welfare variable and weight is equal to 1 if not provided by @randrescastaneda in #48
• Gd lb by @tonyfujs in #45
• Add md_compute_polarization by @Aeilert in #39
• Add interpolation vignette by @Aeilert in #58
• add gd_select_lorenz() function by @tonyfujs in #55
• Input checking by @Aeilert in #61
• Add work on pacakge standardization by @Aeilert in #71
• Add md_fill_gaps() by @Aeilert in #56
• Fill gaps weights by @tonyfujs in #73
• Simplify select lorenz by @tonyfujs in #74
• Fix bug creating vector from dataframe by @randrescastaneda in #81
• Add improved version of fill_gaps() by @Aeilert in #87
• Fix missing dataset by @Aeilert in #93
• Add package authors by @Aeilert in #92
• reduce number of decimals when checking equality to 1 in GD by @randrescastaneda in #90
• Add polarization to dist stats by @Aeilert in #100
• verify missing of survey years in predicted_means by @randrescastaneda in #96
• Fix bugs in gd_clean_data by @Aeilert in #101
• Standardize clean data functions by @Aeilert in #104
• Add gd_compute_dist_stats() by @Aeilert in #105
• Add mean parameter to md_compute_dist_stats() by @Aeilert in #106
• Fix fg create params by @tonyfujs in #110
• Bug fixes by @Aeilert in #111
• fix issue when Watts is failing by @tonyfujs in #115
• Fix unit tests by @Aeilert in #122
• Fix 120 - Gini computation issue by @tonyfujs in #125
• Fix watts by @tonyfujs in #127
• fix error in derive_lb() by @tonyfujs in #129
• fix condition in gd_compute_mld_lb() by @tonyfujs in #131
• Optimize pov stats by @randrescastaneda in #136
• Test gdcomputepipstats2 by @ifeanyi588 in #134
• Remove tibble by @randrescastaneda in #139
• Optimize dist stats by @randrescastaneda in #138
• Handle cases where Watts is numeric(0) by @tonyfujs in #145
• Add prod_* family of function by @tonyfujs in #146
• Add agg by @tonyfujs in #147
• Fill gap agg by @tonyfujs in #150
• add handling of imputed distriubtions for prod_* function family by @tonyfujs in #151
• Function to create synthetic vector by @randrescastaneda in #152
• Testthat3e by @Aeilert in #154
• Prod popshare by @tonyfujs in #156
• Change poverty status threshold by @tonyfujs in #159
• Minor update by @Aeilert in #157
• Revert changes in #157 by @Aeilert in #161
• Styler by @Aeilert in #162
• GH actions by @Aeilert in #163
• Remove order from md_compute_poverty_stats() by @Aeilert in #167
• Use if else instead of ifelse by @Aeilert in #169
• Closes #168 by @Aeilert in #171
• Fix error when headcount is NA by @Aeilert in #170
• Minor optimization of md_compute_poverty_stats by @Aeilert in #181
• Minor optimization of md_infer_poverty_line by @Aeilert in #177
• Minor derive_lq improvment by @Aeilert in #178
• Gd optimizations by @Aeilert in #182
• Regres .lm.fit by @Aeilert in #185
• Optimize regres() by @Aeilert in #183
• fix bug when submitting custom PPP value by @tonyfujs in #187
• Add median to prod functions by @Aeilert in #175
• Prod unit tests by @Aeilert in #180
• Adjust prod fill gaps tests by @tonyfujs in #189
• Use NAreal by @Aeilert in #190
• Add examples by @Aeilert in #192
• Distribution estimator by @tonyfujs in #193
• Fix lorenz selection by @tonyfujs in #195
• fix creation of synth vector for group data. by @randrescastaneda in #196
• md_comptute_gini() improvment by @Aeilert in #197
• MD dist / mld improvment by @Aeilert in #200
• Remove files by @Aeilert in #198
• Remove zzz verbose argument by @Aeilert in #199
• Add lorenz paramter to md_compute_dist_stats() by @Aeilert in #204
• Fix bug by @Aeilert in #203
• Remove old utils functions by @Aeilert in #206
• Simplify documentation by @Aeilert in #207
• Add compute_spl by @Aeilert in #172
• Optimise performance by @shahronak47 in #209
• Fix monotonicity issue by @tonyfujs in #213
• fix value of poverty line with when Beta and quadratic are not normal by @randrescastaneda in #217
• Dev by @tonyfujs in #214

New Contributors

• @Aeilert made their first contribution in #6
• @randrescastaneda made their first contribution in #8
• @tonyfujs made their first contribution in #15
• @ifeanyi588 made their first contribution in #34
• @shahronak47 made their first contribution in #209

Full Changelog: https://github.com/PIP-Technical-Team/wbpip/commits/v0.0.4