climasus4py 0.2.0a2

Fast SUS and climate data workflows for Brazil - Python edition

climasus4py

Fast SUS and climate data workflows for Brazil - Python edition.

Documentation: https://climasus.github.io/climasus4py

Installation

Nota: a versão atual (0.2.0a1) é uma pré-release (alfa). Use a flag --pre para instalá-la. Quando a 0.2.0 estável for publicada, o comando padrão funcionará sem flags adicionais.

Install from PyPI (current pre-release):

pip install --pre climasus4py

Install a specific extra (current pre-release):

pip install --pre "climasus4py[all]"

Install latest from GitHub:

pip install git+https://github.com/climasus/climasus4py.git

Quick Example

import climasus4py as cs

result = cs.sus_pipeline(
    system="SIM-DO",
    uf="SP",
    year=[2021, 2022, 2023],
    lang="en",
    groups=["dengue"],
    time="month",
    geo="state",
)

print(result.df().head())

Step-by-step Example

import climasus4py as cs

# 1. Import and cache
x = cs.sus_data_import("SIM-DO", "SP", [2021, 2022])

# 2. Clean
x = cs.sus_data_clean_encoding(x)

# 3. Standardize
x = cs.sus_data_standardize(x, lang="en")

# 4. Filter
x = cs.sus_filter(x, groups=["dengue"], age_min=0, age_max=80)

# 5. Create variables
x = cs.sus_data_create_variables(x, age_group="who", epi_week=True)

# 6. Aggregate
x = cs.sus_data_aggregate(x, time="month", geo="state")

# 7. Export
cs.sus_export(x, "output/dengue_sp.parquet")

To preserve the original DATASUS .dbc files for audit or reuse, enable the raw cache explicitly:

x = cs.sus_data_import(
    "SINAN-DENGUE",
    "SP",
    2024,
    store_raw=True,
    raw_cache_dir="dados/cache/raw",
)

Shared Metadata

climasus4py consumes metadata from climasus-data (disease groups, dictionaries, UFs, regions, DATASUS FTP sources and SINAN disease codes). This dependency is installed automatically.

You can force update local metadata when needed:

from climasus4py import update_climasus_data
update_climasus_data()

Contributing

• Pull requests and issues are welcome.
• Run tests locally before opening PRs.
• Include a minimal reproducible example for bug reports.

License

MIT