ExtractRINKO (Python Port)

This repo ports the ExtractRINKO Fortran workflow to Python and adds quality-of-life improvements (validation, clearer I/O, chunked output, configurable resampling, and reproducible processing).

Goals

• Reproduce ExtractRINKO results from the Nortek Vector + RINKO EC sensor pipeline.
• Provide a clean, well-tested Python implementation.
• Make inputs/outputs explicit and easier to validate.

Status

Early setup: repo scaffolding and documentation.

Project layout (planned)

• src/: Python implementation
• docs/: reference documents (ignored by git)
• sample_data/: example inputs/outputs (external)

What this does

src/vector_formatter.py processes a converted Nortek Vector data file (text, 18 columns) plus a configuration INI and writes calibrated, resampled output suitable for Aquatic Eddy Covariance (AEC) analysis. It:

• Converts velocities to cm/s, pressure to mbar (with offset), and converts RINKO counts to temperature and oxygen.
• Applies air‑pressure correction (via elevation or measured atmospheric pressure).
• Applies optional time‑based and concentration‑based calibration factors to O2.
• Downscales data to a target frequency by averaging, downsampling, or interpolation.
• Splits output into 30‑minute chunks with timestamped filenames.

Similarities vs ExtractRINKO

• Uses the same core formulas as the Fortran version for temperature, oxygen saturation, density, and oxygen concentration.
• Supports the same calibration inputs and air‑pressure correction options.
• Matches ExtractRINKO outputs (within rounding) when configured equivalently.

Differences vs ExtractRINKO

• Configuration is INI‑based rather than the legacy .def format.
• Output is CSV with labeled header and unit row.
• Output columns are fixed (time, U/V/W, PW, TW, O2, O2_SAT) rather than selectable.
• Output is chunked into 30‑minute files named with a start‑timestamp.
• Resampling is opt‑in and configurable (average, downsample, interpolate).

Configuration

Configuration uses vector_formatter.ini:

• [paths] includes input_file and output_id (default reformatted).
• [sampling] includes hour_start_input, hour_start_output, hour_end_output, input_frequency, output_frequency, resampling_method, and optional start_date/start_time for timestamped filenames.
• [sampling] also includes hour_end_output, which stops processing once the output time exceeds this value.
• When start_date and start_time are provided (e.g., 2025-02-10 and 15:00), each output file name uses the chunk start time formatted as YYYYMMDDTHHMM, and the TIMESTAMP column is written as yyyymmddTHHMM.sssssss.
• [environment] includes pressure_reference, pressure_offset, salinity, elevation, atmospheric_pressure.
• Calibration coefficients are explicit per‑coefficient keys under:
  • [temperature_calibration_coefficients] (AT, BT, CT, DT)
  • [o2_calibration_coefficients] (AO2…HO2)
• Calibration factors live under [calibration_time] and [calibration_concentration].

def_to_ini tool

src/def_to_ini.py converts legacy ExtractRINKO .def files to the new INI schema:

• Reads the .def format (including two/three time‑range lines).
• Writes an INI with explicit coefficient fields and output_id.
• Output defaults to the same path with a .ini extension.

Quick start

Python environment

Create and activate a virtual environment, then install dependencies:

python -m venv .venv
.venv\Scripts\activate
python -m pip install --upgrade pip
pip install -r requirements.txt

Convert a legacy .def file to INI:

python src/def_to_ini.py C:\Users\geryatejina\OneDrive\dev\aec_project\sample_data\sample.def

Basic usage (default = no resampling):

python src/vector_formatter.py C:\Users\geryatejina\OneDrive\dev\aec_project\sample_data\sample.ini

Enable resampling (downsample/average/interpolate):

python src/vector_formatter.py C:\Users\geryatejina\OneDrive\dev\aec_project\sample_data\sample.ini --enable-resample --resample average --target-hz 16

References

See DETAILS.md for a concise breakdown of inputs/outputs and processing steps.