szarr / medusa Goto Github PK

Don't forget to specify dependencies of make_table as well.

Document the rationale and usage of siop_staff to clean the SIOP membership SIDs that the AO provides with the semi-annual data pull.

step_city
step_highestdegree

Both functions have letters outside the standard alphabet and need to be escaped, \u.

This is a single dataframe containing the entire population of data from across the years.

I’m wondering if we can dig into the degree variable a bit more. The dashboard linked below shows ~100 people with missing degrees each year. This is a bit odd because degree is a requirement for SIOP applications. This is a messy variable that we did a lot of recoding for (latest code here: https://github.com/SzarR/medusa/blob/master/R/step_highestdegree.R) because people can list anything (e.g., “ABD”, “certificate in …”). What are we recoding “ABD” into when include_ABD == FALSE? Are those entries getting coded into missing? I believe we decided that ABD should be recoded into Masters, but I’d have to go back and check.

I looked back to our 2020 EB report and we only reported 39 degrees missing that year. Our current 2020 version, that was run through medusa has 108 missing degrees.

Create this as an optional argument for when we need it to run for infographics.

Original file has dates > 2000 < 2022 but the cleaned R file does not.

The content of the warning message:

demo_clean <- medusa::make_demo_data(demo_raw)
Warning messages:
1: Expected 5 pieces. Additional pieces discarded in 1 rows [10197].
2: Expected 5 pieces. Missing pieces filled with NA in 5409 rows [2, 5, 8, 15, 21, 25, 39, 43, 48, 74, 78, 79, 84, 101, 104, 111, 115, 119, 131, 135, ...].

A majority of cases are re-classified as NA when the Priv_Publ function is executed.

• Remove step_field function from the mapping.
• Primary_Career, Occupation values update

Include schools_db as an external database in the package.