Tabular Regression
Scikit-learn
random-forest
scikit-learn
shap
explainable-ai
public-health
air-quality
Instructions to use oukawa/PRISMA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Scikit-learn
How to use oukawa/PRISMA with Scikit-learn:
from huggingface_hub import hf_hub_download import joblib model = joblib.load( hf_hub_download("oukawa/PRISMA", "sklearn_model.joblib") ) # only load pickle files from sources you trust # read more about it here https://skops.readthedocs.io/en/stable/persistence.html - Notebooks
- Google Colab
- Kaggle
|
Download README.md from oukawa/PRISMA: direct link, hf CLI and curl.
- Browser
- Download file 4.86 kB
-
https://huggingface.co/oukawa/PRISMA/resolve/main/README.md
- Command line
-
hf download hf://oukawa/PRISMA/README.md
-
curl -L -o README.md https://huggingface.co/oukawa/PRISMA/resolve/main/README.md
4.86 kB
| license: mit | |
| library_name: sklearn | |
| tags: | |
| - tabular-regression | |
| - random-forest | |
| - scikit-learn | |
| - shap | |
| - explainable-ai | |
| - public-health | |
| - air-quality | |
| # PRISMA: Predicting respiratory hospital admissions in Curitiba and Araucária, Brazil | |
| Code and trained models for the manuscript *Environmental and socioeconomic predictors of respiratory hospital | |
| admissions: An explainable machine learning approach*. | |
| Random forest (RF) models predict hospital admissions for lower respiratory infections (LRI), chronic respiratory | |
| diseases (CRD) and other respiratory conditions (Other) in Curitiba and Araucária (Paraná State, Brazil), 2003–2017, | |
| for four age groups. SHapley Additive exPlanations (SHAP) describe the contribution of each predictor. | |
| - **Temporal models**: weekly admission rates for the city as a whole, from weather, climate indices, air pollutants, | |
| trend and season (24 predictors). | |
| - **Spatial models**: cumulative admissions on a 1-km grid, from socioeconomic variables, population density, road | |
| density and local climate zones (27 predictors). | |
| The models describe associations learned from the data; SHAP values are not causal effects, and correlated predictors | |
| (e.g., temperature and season, or NOx, NO2 and CO) should be interpreted jointly. | |
| ## Files | |
| | File | Content | | |
| |---|---| | |
| | `RF_temporal_{LRI,CRD,Other}_admissions_per_1_million_all_groups.pkl` | Main temporal models (M2, all 24 predictors) | | |
| | `RF_spatial_{LRI,CRD,Other}_admissions_per_1_million_all_groups.pkl` | Main spatial models (S2, all 27 predictors) | | |
| | `RF_temporal_training_and_cross-validation_all_diseases_and_groups.ipynb` | Blocked cross-validation, buffer test, nested models and lags, final models, grid search | | |
| | `RF_spatial_training_and_cross-validation_all_diseases_and_groups.ipynb` | Spatially blocked cross-validation, nested models, final models, grid search | | |
| | `RF_data_processing_temporal_and_spatial.ipynb` | Builds both input matrices following the manuscript (Sections 2.3.1–2.3.4, S1): weekly air quality, weather, climate indices and admission rates; 1-km grid with census, income, GDP, roads, LCZ and admissions | | |
| Each `.pkl` holds a dict with one fitted `RandomForestRegressor` per age group, keyed `{disease}_{group}`, e.g. | |
| `LRI_General`, `LRI_G1`, `LRI_G2`, `LRI_G3`. Age groups: General (all ages), G1 (4 years and under), G2 (5–59 years) | |
| and G3 (60 years and over). Targets are admissions per 1 million inhabitants (per week for the temporal models; | |
| cumulative over 2003–2017 for the spatial models). | |
| All models use 500 trees, a maximum depth of 20, all features at each split, a minimum of two samples to split a | |
| node, one sample per leaf and bootstrap sampling (Table S3). | |
| ## Usage | |
| ```python | |
| import joblib | |
| import pandas as pd | |
| models = joblib.load('RF_temporal_LRI_admissions_per_1_million_all_groups.pkl') | |
| model = models['LRI_G1'] | |
| print(list(model.feature_names_in_)) # predictor columns, in order | |
| # X: DataFrame with these columns -> weekly LRI admissions per 1 million inhabitants (children aged 4 and under) | |
| # pred = model.predict(X[model.feature_names_in_]) | |
| ``` | |
| The models were trained with scikit-learn 1.6.1 and re-saved with scikit-learn 1.8.0 (same trees; predictions agree | |
| with the original files to within 1e-12). Loading them with another version raises a warning, and scikit-learn 1.8 | |
| is recommended. | |
| ## Predictors | |
| **Temporal** (weekly means unless noted): `t_min`, `t_max`, `t_mean`, `t_amp` (°C), `rh` (%), `ws` (m s-1), | |
| `insolation` (h), `evaporation` (mm), `rainfall` (mm, weekly total), `cdd`, `cwd`, `chd`, `ccd` (consecutive dry, wet, | |
| hot and cold days), `julian_day`, `unix_time` (s), `season_fall`, `season_spring`, `season_winter` (summer is the | |
| reference), `so2`, `no2`, `nox`, `o3`, `pm10` (µg m-3) and `co` (mg m-3). | |
| **Spatial** (per 1-km cell): `WHT`, `BLK`, `YEL`, `BRN`, `IND` (% self-reported skin color/race), `LT` (% literate), | |
| `PD` (inhabitants km-2), `PCI` and `GDP` (USD per capita), `RD` (road density, km-1) and `LCZ_1`–`LCZ_10`, | |
| `LCZ_A`–`LCZ_G` (fraction of each local climate zone). | |
| ## Validation | |
| - Temporal: month-per-season blocked cross-validation, six folds, with a 1-week buffer between training and test | |
| weeks (also tested with 0, 2 and 4 weeks). | |
| - Spatial: spatially blocked 5-fold cross-validation with 5 × 5 km blocks. | |
| Scores, nested models, lags and errors by predictor tertile are reported in the Supplementary Material (Tables S2–S8). | |
| ## Data | |
| The input matrices are not included. Hospital admissions are from DataSUS SIH, census data from IBGE (2010), the road | |
| network from OpenStreetMap, local climate zones from the LCZ Generator, and per capita GDP from Kummu et al. (2025). | |
| Air quality and weather data were obtained from IAT and INMET and cannot be shared. The notebooks document the | |
| expected columns. | |
| ## License | |
| MIT | |