q-transform fails on data that used to work
Created by: areeda
This command:
gwpy-plot qtransform --chan L1:GDS-CALIB_STRAIN,reduced --gps 1214870388.0000 --outdir ${HOME}/t --plot 1.00 4.00 16.00 --sample-freq 2048 --frange 0 inf --imin 0.0 --imax 25.5 --search 64 --qrange 4.0 64.0 --cmap viridis
Works with 0.10.1 release but fails with current development branch. Here's a code snippet to demo the problem:
ts2 = TimeSeries.fetch('L1:GDS-CALIB_STRAIN,reduced',1214870388,1214870448)
tmin = ts2.min().value
tmax = ts2.max().value
tdelta = (ts2.max() - ts2.min()).value
asd = ts2.asd().value
amin = asd.min()
print ('TimeSeries min: {:g}, max: {:g}, delta {:g}, asd min {:g}'.format(
tmin, tmax, tdelta, amin))
if amin == 0:
print 'asd min is really zero'
q = ts2.q_transform()
The error message produced is:
invasd = 1. / asd
/usr/share/tomcat/.local/lib/python2.7/site-packages/gwpy/timeseries/timeseries.py:1511: RuntimeWarning: invalid value encountered in multiply
out.value[x:y] += npfft.irfft(in_.fft().value * invasd)
/usr/share/tomcat/.local/lib/python2.7/site-packages/numpy/lib/function_base.py:4033: RuntimeWarning: Invalid value encountered in median
r = func(a, **kwargs)
/usr/share/tomcat/.local/lib/python2.7/site-packages/numpy/core/_methods.py:26: RuntimeWarning: invalid value encountered in reduce
return umr_maximum(a, axis, None, out, keepdims)
---------------------------------------------------------------------------
UnboundLocalError Traceback (most recent call last)
<ipython-input-10-5bb307ef33cc> in <module>()
----> 1 q = ts2.q_transform()
/usr/share/tomcat/.local/lib/python2.7/site-packages/gwpy/timeseries/timeseries.pyc in q_transform(self, qrange, frange, gps, search, tres, fres, norm, outseg, whiten, **asd_kw)
1726 # (Q, frequency) `TimeSeries` to have the same time resolution
1727 nx = int(abs(Segment(*outseg)) / tres)
-> 1728 ny = frequencies.size
1729 out = Spectrogram(numpy.zeros((nx, ny)), x0=outseg[0], dx=tres,
1730 frequencies=frequencies)
UnboundLocalError: local variable 'frequencies' referenced before assignment
This is a collection of plots to help understand the data: https://ldvw.ligo.caltech.edu/ldvw/view?act=imagehistory&strt=0&pageNum=1&usrSel=Joseph+Areeda+%283357%29&group=Zero+ASD&size=original&op_group=Favorites&newGrpName=%3Cnew+group+name%3E