Job request: 24575

version: '3.0'

expectations:

  population_size: 5000

actions:

  ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 
  ## DO NOT EDIT project.yaml DIRECTLY 
  ## This file is created by create_project_actions.R 
  ## Edit and run create_project_actions.R to update the project.yaml 
  ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 
  ## Define study dates 

  study_dates:
    run: r:latest analysis/study_dates.R
    outputs:
      highly_sensitive:
        study_dates_json: output/study_dates.json

  ## Generate dates for all cohorts 

  generate_dates:
    run: ehrql:v1 generate-dataset analysis/dataset_definition/dataset_definition_dates.py
      --output output/dataset_definition/index_dates.csv.gz
    needs:
    - study_dates
    outputs:
      highly_sensitive:
        dataset: output/dataset_definition/index_dates.csv.gz

  ## Generate input_prevax 

  generate_input_prevax:
    run: ehrql:v1 generate-dataset analysis/dataset_definition/dataset_definition_prevax.py
      --output output/dataset_definition/input_prevax.csv.gz
    needs:
    - generate_dates
    outputs:
      highly_sensitive:
        cohort: output/dataset_definition/input_prevax.csv.gz

  ## Generate input_prevax_clean, with describe = TRUE 

  generate_input_prevax_clean:
    run: r:latest analysis/dataset_clean/dataset_clean.R prevax TRUE
    needs:
    - study_dates
    - generate_input_prevax
    outputs:
      moderately_sensitive:
        describe_raw: output/describe/prevax_raw.txt
        describe_venn: output/describe/prevax_venn.txt
        describe_preprocessed: output/describe/prevax_preprocessed.txt
        flow: output/dataset_clean/flow-cohort_prevax.csv
        flow_midpoint6: output/dataset_clean/flow-cohort_prevax-midpoint6.csv
      highly_sensitive:
        venn: output/dataset_clean/venn-cohort_prevax.rds
        cohort_clean: output/dataset_clean/input_prevax_clean.rds

  ## Run models 

  make_model_input-cohort_prevax-main-ami:
    run: r:latest analysis/model/make_model_input.R cohort_prevax-main-ami
    needs:
    - generate_input_prevax_clean
    outputs:
      highly_sensitive:
        model_input: output/model/model_input-cohort_prevax-main-ami.rds

  cox_ipw-cohort_prevax-main-ami:
    run: cox-ipw:v0.0.37 --df_input=model/model_input-cohort_prevax-main-ami.rds --ipw=TRUE
      --exposure=exp_date --outcome=out_date --strata=strat_cat_region --covariate_sex=cov_cat_sex
      --covariate_age=cov_num_age --covariate_other=cov_cat_ethnicity;cov_cat_imd;cov_num_consrate2019;cov_bin_hcworker;cov_cat_smoking;cov_bin_carehome;cov_bin_obesity;cov_bin_ami;cov_bin_dementia;cov_bin_liver_disease;cov_bin_ckd;cov_bin_cancer;cov_bin_hypertension;cov_bin_diabetes;cov_bin_depression;cov_bin_copd;cov_bin_stroke_all;cov_bin_other_ae;cov_bin_vte;cov_bin_hf;cov_bin_angina;cov_bin_lipidmed;cov_bin_antiplatelet;cov_bin_anticoagulant;cov_bin_cocp;cov_bin_hrt
      --cox_start=index_date --cox_stop=end_date_outcome --study_start=2020-01-01
      --study_stop=2024-04-30 --cut_points=1;7;14;28;56;84;183;365;730;1065;1582 --controls_per_case=20
      --total_event_threshold=50 --episode_event_threshold=5 --covariate_threshold=5
      --age_spline=TRUE --save_analysis_ready=FALSE --run_analysis=TRUE --df_output=model/model_output-cohort_prevax-main-ami.csv
    needs:
    - make_model_input-cohort_prevax-main-ami
    outputs:
      moderately_sensitive:
        model_output: output/model/model_output-cohort_prevax-main-ami.csv

  make_model_input-cohort_prevax-main-stroke_sahhs:
    run: r:latest analysis/model/make_model_input.R cohort_prevax-main-stroke_sahhs
    needs:
    - generate_input_prevax_clean
    outputs:
      highly_sensitive:
        model_input: output/model/model_input-cohort_prevax-main-stroke_sahhs.rds

  cox_ipw-cohort_prevax-main-stroke_sahhs:
    run: cox-ipw:v0.0.37 --df_input=model/model_input-cohort_prevax-main-stroke_sahhs.rds
      --ipw=TRUE --exposure=exp_date --outcome=out_date --strata=strat_cat_region
      --covariate_sex=cov_cat_sex --covariate_age=cov_num_age --covariate_other=cov_cat_ethnicity;cov_cat_imd;cov_num_consrate2019;cov_bin_hcworker;cov_cat_smoking;cov_bin_carehome;cov_bin_obesity;cov_bin_ami;cov_bin_dementia;cov_bin_liver_disease;cov_bin_ckd;cov_bin_cancer;cov_bin_hypertension;cov_bin_diabetes;cov_bin_depression;cov_bin_copd;cov_bin_stroke_all;cov_bin_other_ae;cov_bin_vte;cov_bin_hf;cov_bin_angina;cov_bin_lipidmed;cov_bin_antiplatelet;cov_bin_anticoagulant;cov_bin_cocp;cov_bin_hrt
      --cox_start=index_date --cox_stop=end_date_outcome --study_start=2020-01-01
      --study_stop=2024-04-30 --cut_points=1;7;14;28;56;84;183;365;730;1065;1582 --controls_per_case=20
      --total_event_threshold=50 --episode_event_threshold=5 --covariate_threshold=5
      --age_spline=TRUE --save_analysis_ready=FALSE --run_analysis=TRUE --df_output=model/model_output-cohort_prevax-main-stroke_sahhs.csv
    needs:
    - make_model_input-cohort_prevax-main-stroke_sahhs
    outputs:
      moderately_sensitive:
        model_output: output/model/model_output-cohort_prevax-main-stroke_sahhs.csv