This python package defines a reader and helper methods to handle data exported from the Luna ODiSI 6000 optical measuring system. It allows for an easier retrieval of data corresponding to each segment, as well as the possibility to interpolate the results based on additional measurements, such as experimental load. Doing this manually requires some amount of python code, which can be avoided by using this package.

Install as usual:

pip install python-odisi

The library can be used to read files in the following manner:

from odisi import read_tsv

d = read_tsv("data_gages.tsv")

# List all gages
gages = d.gages
# List all segments
segments = d.segments
# Get the data for a specific gage, e.g. with the label 'A'
d_gage = d.gage("A")
# Get the data for a specific segment, e.g. with the label 'Seg-1'
d_seg, x_seg = d.segment("Seg-1")

The package allows to easily interpolate an external signal (e.g. the load during the test). For this, two strategies can be followed:

import polars as pl

load = pl.read_csv("load_data.csv")
# Assume that the timestamp is in the column 'time'
d.interpolate(load.select(pl.col("time")))

Then you should be able to plot your data against the measured load:

import matplotlib.pyplot as plt

d_gage = d.gage("A")
# Assume that the load data is in column 'load'
a_load = load.select(pl.col("load")).to_series()

plt.plot(d_gage, a_load)

import polars as pl

load = pl.read_csv("load_data.csv")
# Assume that the timestamp is in the column 'time'
new_load = d.interpolate_signal(data=load, time="time")

Then you should be able to plot your data against the measured load:

import matplotlib.pyplot as plt

d_gage = d.gage("A")
# Assume that the load data is in column 'load'
a_load = new_load.select(pl.col("load")).to_series()

plt.plot(d_gage, a_load)

In both cases it is assumed that the timestamps from both files are synchronized, i.e. that both measuring computers have synchronized clocks.

It is probable that the measurements from both data sources (ODiSI and additional system) were started at different times. This produces some annoyances during the processing of the data due to the mismatch in datapoints. To remedy this, the option clip=True can be passed to both interpolation methods (interpolate(...) and interpolate_signal(...)), which will clip the data to the common time interval between both signals.

import polars as pl

load = pl.read_csv("load_data.csv")
# Assume that the timestamp is in the column 'time'
d.interpolate(load.select(pl.col("time")), clip=True)

The data of all segments can be exported to individual csv-files with the following code:

d.export_segments_csv(prefix="my_experiment", path="data_folder")

The package includes a test suite which should be run with pytest:

poetry run pytest

@software{Tapia_2023,
    author = {Tapia Camú, Cristóbal},
    title = {{python-odisi: Import data generated by the Luna ODiSI System}},
    url = {https://github.com/cristobaltapia/python-odisi},
    version = {v0.3},
    year = {2023},
}