feat: edvise schema school data audit

[nm3224]

feat: edvise schema school data audit

• NOTE: ensure PR feat: adding bias variable check to data audit #86 and PR feat: adding custom school functions to repo #88 are merged before merging this or it will create several merge conflicts.
• The structure for the custom cohort and course file standardizers are as follows:

Edvise Cohort Standardizer:

• Operates similarly to PDP's cohort standardizer.
• Logs credential types and enrollment types
• Log high values of NAs
• Logs missing bias variables
• Logs grade distribution
• Logs top majors
• Replaces NA fields with "N" in pell and first gen columns
• Finds and logs duplicates on primary keys: ["student_id", "cohort_term"]
  - Runs drop readmits func; checks again for remaining duplicates, if yes, runs keep_earlier_record func
• Drops unused, unpopulated bias columns

Edvise Course Standardizer:

• Operates similarly to PDP's course standardizer.
• Strip trailing decimals
• Log high values of NAs
• Log grade distribution
• Finds and logs duplicates on primary keys: ["student_id", "term", "course_subject", "course_num"]
  - Runs handling duplicates func (schema_type="es")
• Runs check_pf_grade_consistency func
• Runs validate_credit_consistency func
• Runs assign_numeric_grade func

Logging cohort years + terms, academic years + terms (this can happen in the actual script, similar to how it is for PDP)

Questions

• I'm not sure if the approach right now with dropping re admits and keeping the earlier record is the best way to de-dupe duplicates in the cohort file; i'm finding for some schools, duplicates are related to dual major or dual program students. How might we address these?
• We still need to edit the handling duplicates function in the course file to fit a default standard for our edvise schema schools.

[kaylawilding]

Edited to include the edivse schema data audit:

• pulled the alike code into a data audit class in the backend
• now the edvise and pdp scripts simply call that and do cli