matbbastos / epw-analysis

Processing and exporting data from EPW files into other formats.

ARCHIVED

This repository has been archived. Another repository now contains most of the content and functionality that was once here. The new repository will receive any updates and fixes, while this one will not.

Link to NEW REPOSITORY: https://github.com/igorcmvaz/future-EPW-analysis

Old README

EPW Analysis

A script to process multiple EPW files into a single (Apache) Parquet file, taking only the desired data columns and informational fields.

How to use

git clone [email protected]:Matbbastos/epw-analysis.git
python -m pip install -r requirements.txt

python -m merge_files_into_parquet.py -h
python -m merge_files_into_parquet.py sample_input --csv

Confort Models

Calculated variables are introduced in the files based on the models from pythermalcomfort. For more information, visit their documentation site. The models are included in the exported files in an 'opt-out' format, so that passing the -s or --strict option will prevent pythermalcomfort from being imported, skip computation of the models and exclude corresponding entries from the output files.

Note: Bear in mind that merely importing pythermalcomfort comes with a significant overhead for the script, and the computation of the values for comfort models are executed in each iteration, also noticeably impacting overall performance. For that reason, the imports are conditional and only executed if the models are in fact desired (no --strict option).

An extra option (-l or --limit-utci) was added to the script, and it is used to configure the flag of similar name in UTCI model, as detailed in the corresponding documentation (check parameter limit_inputs). If this option is present, this flag is used when computing UTCI models and the Wind Speed values from EPW files are saturated using the lower/upper bounds from the model (as per documentation), to ensure they are within the working range. Default is to not limit the inputs and use extrapolation from the UTCI model itself (rather than saturating the values).

Commits

When committing to this repository, following convention is advised:

• chore: regular maintenance unrelated to source code (dependencies, config, etc)
• docs: updates to any documentation
• feat: new features
• fix: bug fixes
• ref: refactored code (no new feature or bug fix)
• revert: reverts on previous commits
• test: updates to tests

For further reference on writing good commit messages, see Conventional Commits.