Visualization Deep Dive

SegmentTable provides rich plotting features, specifically optimized for comparing spectra over multiple segments via overlay_spectra().

[ ]:

import numpy as np
from gwpy.segments import Segment
from gwexpy.table import SegmentTable
from gwpy.frequencyseries import FrequencySeries

def make_fs(i):
    f = np.linspace(1, 32, 256)
    data = (1.0/(f**1.5)) * (1.0 + i*0.1)
    return FrequencySeries(data, frequencies=f)

segs = [Segment(i*100, i*100+100) for i in range(10)]
st = SegmentTable.from_segments(segs, snr=np.random.uniform(5, 20, 10))
st.add_series_column("asd", data=[make_fs(i) for i in range(10)], kind="frequencyseries")

# 1. Overlay spectra graded by start time (default)
plot = st.overlay_spectra("asd", color_by="t0")
plot.show() # NOTE: Normally use plot in notebooks

Color Grading by Meta Column

You can color individual lines based on any numeric meta column (e.g., SNR).

[ ]:

plot = st.overlay_spectra("asd", color_by="snr", cmap="plasma")
plot

Overview Layouts

Use segments() to see the temporal layout of your table.

[ ]:

st.segments(color="snr")