Add malaria phase 2 modeling script with data processing and model comparisons

- Implemented data loading from parquet files for malaria forecasting
- Normalized malaria metrics by population
- Filtered data based on mortality thresholds and location hierarchy
- Centered various covariates by location
- Developed linear models to analyze the relationship between malaria prevalence and various predictors
- Created functions for model comparison and result parsing
- Visualized model results using beeswarm plots and R-squared metrics
- Included GAM modeling for non-linear relationships

Author: bcreiner

poetry.lock

@@ -15,6 +15,10 @@ dependencies = [
     "rra-tools (>=1.0.25,<2.0.0)",
     "rasterio (>=1.4.3,<2.0.0)",
     "ipykernel (>=6.29.5,<7.0.0)",
+    "xarray (>=2025.4.0,<2026.0.0)",
+    "scipy (>=1.15.2,<2.0.0)",
+    "netcdf4 (>=1.7.2,<2.0.0)",
+    "rasterra (>=0.6.2,<0.7.0)",
 ]
 
 [project.urls]

pyproject.toml

@@ -1,5 +1,5 @@
 # I need for each covariate
-# covariate: covariate name
+# covariate_name: covariate name
 # covariate_resolution: covariate raster resolution in degrees
 # years: list of years
 # synoptic: T/F
@@ -9,20 +9,102 @@
 
 COVARIATE_DICT:
   elevation:
-    covariate: "elevation"
+    covariate_name: "elevation"
     covariate_resolution_numerator: 5
     covariate_resolution_denominator: 120
     year_start: 2020
     year_end: 2020
     synoptic: true
     cc_sensitive: false
     summary_statistic: "mean"
+    path: "/snfs1/WORK/11_geospatial/01_covariates/00_MBG_STANDARD/elevation/mean/2019_06_10/synoptic/elevation_mean_synoptic.tif"
   malaria_pfpr:
-    covariate: "malaria_pfpr"
+    covariate_name: "malaria_pfpr"
     covariate_resolution_numerator: 5
     covariate_resolution_denominator: 120
     year_start: 2000
-    year_end: 2023 
+    year_end: 2022 
     synoptic: false
     cc_sensitive: false
     summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-pfpr/2024_GBD2023_Global_PfPR_{year}.tif"
+  malaria_pf_inc_count:
+    covariate_name: "malaria_pf_inc_count"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2022
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-incidence-count/2024_GBD2023_Global_Pf_Incidence_Count_{year}.tif"
+  malaria_pf_inc_rate:
+    covariate_name: "malaria_pf_inc_rate"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2022
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-incidence-rate/2024_GBD2023_Global_Pf_Incidence_Rate_{year}.tif"
+  malaria_pf_mort_count:
+    covariate_name: "malaria_pf_mort_count"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2022
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-mortality-count/2024_GBD2023_Global_Pf_Mortality_Count_{year}.tif"
+  malaria_pf_mort_rate:
+    covariate_name: "malaria_pf_mort_rate"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2022
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-mortality-rate/2024_GBD2023_Global_Pf_Mortality_Rate_{year}.tif"
+  malaria_irs:
+    covariate_name: "malaria_irs"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2020
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-irs/202106_Africa_Indoor_Residual_Spraying_Coverage_{year}.tif"
+  malaria_itn:
+    covariate_name: "malaria_itn"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2020
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-itn/202106_Africa_Insecticide_Treated_Net_Use_{year}.tiff"
+  malaria_anti_viral:
+    covariate_name: "malaria_anti_viral"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2000
+    year_end: 2020
+    synoptic: false
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/malaria-anti/202106_Global_Antimalarial_Effective_Treatment_{year}.tif"
+  dengue_suitability:
+    covariate_name: "dengue_suitability"
+    covariate_resolution_numerator: 5
+    covariate_resolution_denominator: 120
+    year_start: 2015
+    year_end: 2015
+    synoptic: true
+    cc_sensitive: false
+    summary_statistic: "mean"
+    path: "/mnt/team/rapidresponse/pub/malaria-denv/data/02-processed-data/dengue-suit/DEN_rangemap_2023_05_31.tif"

src/idd_forecast_mbp/COVARIATE_DICT.yaml

@@ -25,12 +25,13 @@ def parse_yaml_dictionary(covariate: str) -> dict:
 
     # Build the return dict dynamically
     result = {
-        "covariate": covariate_entry['covariate'],
+        "covariate_name": covariate_entry['covariate_name'],
         "covariate_resolution": covariate_resolution,
         "years": years,
         "synoptic": covariate_entry['synoptic'],
         "cc_sensitive": covariate_entry['cc_sensitive'],
-        "summary_statistic": covariate_entry['summary_statistic']
+        "summary_statistic": covariate_entry['summary_statistic'],
+        "path": covariate_entry['path'],
     }
 
     return result