feat: loanbook_demo and abcd_demo have no non-NA values for lei_direct_loantaker or lei respectively

[jdhoffa]

Expected behaviour:
At least one non-NA value for the lei relevant column.

Observed behaviour:

library(r2dii.data)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

nrow(dplyr::filter(loanbook_demo, !is.na(lei_direct_loantaker)))
#> [1] 0

nrow(dplyr::filter(abcd_demo, !is.na(lei)))
#> [1] 0

^{Created on 2024-03-06 with reprex v2.1.0}

[jdhoffa]

Weirdly, the lei columns seem to have been removed purposefully in #316

@cjyetman any idea why this was done? The LEI calculating functionality produces entirely fake LEIs and adds them to a subset of entries. It is particularly critical now that RMI-PACTA/r2dii.match#460 is merged.

[cjyetman]

Weirdly, the lei columns seem to have been removed purposefully in #316

@cjyetman any idea why this was done? The LEI calculating functionality produces entirely fake LEIs and adds them to a subset of entries. It is particularly critical now that RMI-PACTA/r2dii.match#460 is merged.

Pretty sure I just started from https://github.com/RMI-PACTA/r2dii.data/blob/main/data-raw/loanbook_demo.csv and modified things to make it all fake... I guess I did not know to add LEIs to it.

[jdhoffa]

But that's the thing, the LEI's were already calculated, and then they were explicitly removed in this line:

r2dii.data/data-raw/abcd_demo.R

Lines 197 to 198 in e57b1d3

	# remove LEIs
	abcd_demo <- mutate(abcd_demo, lei = NA)

I was just curious if this was done intentionally/ why. But it sounds like there isn't a reason for it?

[cjyetman]

But that's the thing, the LEI's were already calculated, and then they were explicitly removed in this line:

r2dii.data/data-raw/abcd_demo.R

Lines 197 to 198 in e57b1d3

	# remove LEIs
	abcd_demo <- mutate(abcd_demo, lei = NA)

I was just curious if this was done intentionally/ why. But it sounds like there isn't a reason for it?

hmmm.... now that you point that out, I vaguely remember this being a last minute request from George (to avoid sharing anything sensitive?). Maybe @jacobvjk or George would remember?

[jdhoffa]

Yeah, I had a feeling it was that.

In any case, there was never anything sensitive in the data, since those values are calculated programmatically/ randomly earlier on:

r2dii.data/data-raw/abcd_demo.R

Lines 9 to 37 in e57b1d3

	generate_lei <- function(id) {
	# function to generate random but reproducible LEIs
	# 4 characters, 2 zeroes, 12 characters, 2 check digits
	alpha_num <- c(0:9, LETTERS)
	withr::with_seed(
	id,
	{
	four <- do.call(
	paste0,
	replicate(4, sample(0:9, 1, TRUE), FALSE)
	)
	twelve <- do.call(
	paste0,
	replicate(12, sample(alpha_num, 1, TRUE), FALSE)
	)
	two <- do.call(
	paste0,
	replicate(2, sample(0:9, 1, TRUE), FALSE)
	)
	}
	)
	paste0(four, "00", twelve, two)
	}
	vgenerate_lei <- Vectorize(generate_lei)