Job request: 8278

version: '3.0'

expectations:
  population_size: 10000

actions:

  ####################
  # Cohort Generation
  ####################

  # Since this runs on everyone, we can reuse for both studies 
  generate_study_population_ethnicity:
    run: cohortextractor:latest generate_cohort --study-definition study_definition_ethnicity --output-format=csv.gz
    outputs:
      highly_sensitive:
        cohort: output/input_ethnicity.csv.gz

  # Generate depression register, dep003, depression and prescription variables by month
  generate_study_population:
    run: cohortextractor:latest generate_cohort --study-definition study_definition --index-date-range "2019-01-01 to 2022-01-01 by month" --output-format=csv.gz
    outputs:
      highly_sensitive:
        cohort: output/input_*.csv.gz

  ####################
  # Join ethnicity to all generated input files
  # Efficiency fix https://github.com/opensafely/research-template
  # BUT BEWARE STALE DATA
  ###################

  join_cohorts:
    run: >
      cohort-joiner:v0.0.18
        --lhs output/input_*.csv.gz
        --rhs output/input_ethnicity.csv.gz
        --output-dir output/joined
    needs: [generate_study_population_ethnicity, generate_study_population]
    outputs:
      highly_sensitive:
        cohort: output/joined/input_*.csv.gz

  ####################
  # Measures 
  ####################

  # Output the summary values by date
  generate_measures:
      run: cohortextractor:latest generate_measures --study-definition study_definition --output-dir=output/joined
      needs: [join_cohorts]
      outputs:
        moderately_sensitive:
          # Only output the single summary file
          measure_csv: output/joined/measure_*_rate.csv

  #############################
  # Plotting
  #############################
  generate_deciles_charts:
    run: >
            deciles-charts:v0.0.15
            --input-files output/joined/measure_*_practice_rate.csv
            --output-dir output/joined
    config:
      show_outer_percentiles: false
      tables:
        output: true
      charts:
        output: true
    needs: [generate_measures]
    outputs:
      moderately_sensitive:
        cohort: output/joined/deciles_*_*.*

  generate_groups:
    run: >
            python:latest python analysis/group_charts.py
            --input-files output/joined/measure_*.csv
            --output-dir output/joined
            --date-lines "2019-03-31" "2020-03-31" "2021-03-31"
            --scale "rate"
    needs: [generate_measures]
    outputs:
      moderately_sensitive:
        cohort: output/joined/group_chart_*.png

  # TODO: overwriting plots for small subset the require percentage
  generate_percentage_groups:
    run: >
            python:latest python analysis/group_charts.py
            --input-files output/joined/measure_dep003*.csv
            --output-dir output/joined
            --date-lines "2019-03-31" "2020-03-31" "2021-03-31"
            --scale "percentage"
    needs: [generate_measures]
    outputs:
      moderately_sensitive:
        cohort: output/joined/group_chart_dep003*.png


  generate_table1:
    run: >
            python:latest python analysis/table1.py
            --input-dir output/joined
            --output-dir output/joined
            --measure-attribute "antidepressant_any"
    needs: [generate_measures]
    outputs:
      moderately_sensitive:
        cohort: output/joined/table1.csv

  #############################
  # Display
  #############################
  generate_report:
    run: >
            python:latest python analysis/report.py
            --input-dir output/joined
            --output-dir output/joined
            --resource-dir analysis/resources
    needs: [generate_deciles_charts, generate_groups, generate_percentage_groups]
    outputs:
      moderately_sensitive:
        cohort: output/joined/report.html