Skip to main content

Loading Data

varistar.catalog has one loader per survey format. TimeSeries.load_data_from_file auto-dispatches based on file extension; the catalog modules can also be called directly when you need more control.

OGLE-style .dat files​

from varistar import TimeSeries

ts = TimeSeries(magnitude="mag I", time_scale="HJD")
ts.load_data_from_file("data/OGLE-BLG-ECL-000001.dat")  # -> catalog.ogle.load_dat
ts.summary()

TESS​

from varistar.catalog.tess import load_from_tic, load_fits

df = load_from_tic(tic_id=123456789)        # via lightkurve
df = load_fits("data/tess_sector_42.fits")  # from a local FITS file

Gaia DR3 epoch photometry​

from varistar.catalog.gaiadr3 import load_csv

df = load_csv("data/gaia_dr3_epoch_photometry.csv")

Generic CSV / your own arrays​

For anything else — a lab CSV, a pipeline export, or data you already have in memory — use the generic loader or build a DataFrame directly and hand it to load_data_from_df. This is also what every other page in this docs instance uses under the hood:

from varistar import TimeSeries
from _synthetic import make_sinusoidal_lightcurve

df = make_sinusoidal_lightcurve()

ts = TimeSeries(magnitude="mag I", time_scale="HJD")
ts.load_data_from_df(df, data_id="synthetic_001")
ts.summary()
[synthetic_001]  n=400  baseline=118.2 d  <mag>=14.991  amp=0.579

Starting from three plain numpy arrays instead of a DataFrame works the same way via catalog.generic.from_arrays:

import numpy as np
from varistar.catalog.generic import from_arrays

t = np.asarray(df["hjd"])
y = np.asarray(df["mag_i"])
e = np.asarray(df["m_error"])

pl_df = from_arrays(t, y, e, col_names=["hjd", "mag_i", "m_error"])

ts2 = TimeSeries(magnitude="mag I", time_scale="HJD")
ts2.load_data_from_df(pl_df, data_id="from_arrays_demo")
len(ts2)
400

from_arrays defaults to column names ["hjd", "mag", "err"] — pass col_names explicitly if you want them to match a TimeSeries built with custom colnames.

Next: Cleaning & Statistics.