Job request: 25884

version: '4.0'

actions:
  # generate_dataset:
  #   run: ehrql:v1 generate-dataset analysis/dataset_definition_patients.py --output output/dataset_patients.csv.gz
  #   outputs:
  #     highly_sensitive:
  #       dataset: output/dataset_patients.csv.gz

  generate_dataset_preg_2024_11_01:
    run: ehrql:v1 generate-dataset analysis/dataset_definition_pregnancy.py --output output/dataset_preg_2024_11_01.csv.gz -- --start_date 2024-11-01
    outputs:
      highly_sensitive:
        dataset: output/dataset_preg_2024_11_01.csv.gz

  generate_dataset_preg_2025_11_01:
    run: ehrql:v1 generate-dataset analysis/dataset_definition_pregnancy.py --output output/dataset_preg_2025_11_01.csv.gz -- --start_date 2025-11-01
    outputs:
      highly_sensitive:
        dataset: output/dataset_preg_2025_11_01.csv.gz

  generate_dataset_preg_measures:
    run: ehrql:v1 generate-measures analysis/dataset_definition_pregnancy_measures.py --output output/measures/:csv
    outputs:
      moderately_sensitive:
        dataset: output/measures/*.csv

  analyse_preg_measures:
    run: python:v2 analysis/pregnancy_checks.py --output output/measures/pregnancy_checks.csv --input output/measures/*.csv
    needs: 
      [generate_dataset_preg_measures]
    outputs:
      moderately_sensitive:
        dataset: output/measures/*by_year.csv
        charts: output/measures/*by_year.png


  # generate_dataset_practices:
  #   run: ehrql:v1 generate-dataset analysis/dataset_definition_practices.py --output output/dataset_practices.csv.gz
  #   outputs:
  #     highly_sensitive:
  #       dataset: output/dataset_practices.csv.gz
  
  generate_patient_dataset_2024_02_01:
    run: ehrql:v1 generate-dataset analysis/dataset_definition_patients.py --dummy-tables dummy-folder --output output/dataset_patients_2024-02-01.csv.gz -- --start_date 2024-02-01
    outputs:
      highly_sensitive:
        dataset: output/dataset_patients_2024-02-01.csv.gz
  generate_patient_dataset_2024_03_01:
    run: ehrql:v1 generate-dataset analysis/dataset_definition_patients.py --dummy-tables dummy-folder --output output/dataset_patients_2024-03-01.csv.gz -- --start_date 2024-03-01
    outputs:
      highly_sensitive:
        dataset: output/dataset_patients_2024-03-01.csv.gz

  # combine monthly tables
  combine_monthly_patient_gz:
    run: python:v2 analysis/preprocess_combine_gz.py --output output/dataset_patients_combined.csv.gz
    needs: [generate_patient_dataset_2024_02_01, generate_patient_dataset_2024_03_01]
    outputs:
      highly_sensitive:
        dataset: output/dataset_patients_combined.csv.gz

  # patient-level measures for sense checking; not final measures for analysis
  # local testing with dummy tables
  generate_patient_measures:
    run: ehrql:v1 generate-measures analysis/measures_patient.py --dummy-tables dummy-folder --output output/patient_measures.csv
    outputs:
      moderately_sensitive:
        measure_tables: output/patient_measures.csv
  
  
  # practice-level data step 1: run patient dataset measures 
  generate_patient_measures_for_practices:
    run: ehrql:v1 generate-measures analysis/measures_practice.py --dummy-tables dummy-folder --output output/practice_measures.csv
    outputs:
      highly_sensitive:
        measure_tables: output/practice_measures.csv
  
  # practice-level data step 2: aggregate patient measures to practice-level data
  process_measures_to_practice_dataset:
    run: python:v2 analysis/process_measures_to_practice_dataset.py --output output/practice_level_data.csv
    needs: [generate_patient_measures_for_practices]
    outputs:
      moderately_sensitive:
        practice_level_data: output/practice_level_data.csv
  
  # practice-level data step 3: extract info from practice-level data
  process_practice_dataset:
    run: python:v2 analysis/process_practice_dataset.py --output output/practice_summary_by_stp.csv
    needs: [process_measures_to_practice_dataset]
    outputs:
      moderately_sensitive:
        practice_level_data: output/practice_summary_by_stp.csv

# https://actions.opensafely.org/actions/safetab/v3.0.2/
# safetabs: create two-way tables for the specified variables; redact results for small numbers
  generate_safetabs:
    run: safetab:v3.0.2 output/dataset_patients_combined.csv.gz
    needs: [combine_monthly_patient_gz]
    config:
      output_path: output
      redaction_limit: 2
      tables:
        two_way:
          tab_type: 2-way
          variables:
            - sex
            - region
    outputs:
      moderately_sensitive:
        safetab_log: output/dataset_patients_combined.csv_tables/table_log.txt
        safetab_two_way: output/dataset_patients_combined.csv_tables/two_way/*.md

# https://actions.opensafely.org/actions/cohort-report/v3.0.0/
# generate cohort report; require all variable_types
  # generate_cohort_report:
  #   run: cohort-report:v3.0.0 output/dataset_patients_combined.csv.gz
  #   needs: [combine_monthly_patient_gz]
  #   config:
  #     variable_types:
  #       start_date: date
  #       index_date: date
  #       registered_start: categorical
  #       registered_index: categorical
  #       alive: categorical
  #       sex: categorical
  #       age: int
  #       date_of_birth: date
  #       imd: categorical
  #       ethnicity: categorical
  #       region: categorical
  #   outputs:
  #     moderately_sensitive:
  #       reports: cohort_reports_outputs/descriptives_dataset_patients_combined.csv.html
  #       figures: cohort_reports_outputs/*.png

# https://actions.opensafely.org/actions/dataset-report/v0.0.29/
# dataset report
  generate_dataset_report:
    run: >
      dataset-report:v0.0.29
        --input-files output/dataset_patients_combined.csv.gz
        --output-dir output
    needs: [combine_monthly_patient_gz]
    outputs:
      moderately_sensitive:
        dataset_report: output/dataset_patients_combined.html