Skip to content

AGENTS.md

Latest commit

 

History

History
67 lines (47 loc) · 2.18 KB

AGENTS.md

File metadata and controls

67 lines (47 loc) · 2.18 KB

AGENTS.md

Project summary

HLAVO is a hydrology modeling system focused on predicting groundwater table dynamics from meteorological inputs and soil moisture profile measurements. The repo combines data ingress pipelines, surface modeling (Richards/ParFlow) with Kalman-filter assimilation, and deep vadose zone model generation from GIS into MODFLOW 6 inputs, with runs/configs under runs/.

Project summary (short)

Hydrology modeling stack for the Uhelná locality: ingest meteo + soil moisture, assimilate in surface models, and build deep vadose zone MODFLOW 6 models from GIS sources.

CODEX Ignore Folders

  • DeepSensor
  • microchip_SW
  • notebooks
  • _*

CODEX Readonly Folders

  • dev
  • hlavo/deep_model/GIS

hlavo/ Project Source Structure

  • composed/ - communication of deep_model and kalman (surface model) in a parallel run
  • deep_model/ - model of the deep vadoze zone
  • ingress/ - data ingress tools and services
  • kalman/ - surface model assimilation (using profile measurements)
  • soil_parflow/ - ParFlow based surface model (Richards)
  • soil_py/ - Pure Python minimalistic Richards' solver.

HLAVO Project Guidelines

This is subproject of HLAVO repository living in deep_model/gpt_zone. Codex run within docker container

Coding rules

  • Best code, is no code!
  • prefere functional style with poor functions; idealy do not change objects after construction, all methods do calculations only reading the data in the class
  • prefere high level code: numpy, pandas, xarray instead loops and native python sturctures (lists, dicts)
  • use logging
  • Use logging for debug outputs.
  • use pathlib
  • use attrs for dataclasses
  • use attrs staticmethod/classmethod technique to construct from other data then is stored in the dataclass
  • Be defensive, with strong checks, but only for the user input data.
  • Do just basic asserts for consistency for function inputs.
  • Can add more asserts if needed during debugging.
  • NEVER resolve test errors by try blocks

How to verify your changes

  • Use tests/run script to run pytest as it redirects merged stdin + stderr into pytest.log otherwise you may mis the actual output, which breaks your feedback loop.