Running ./specphase generates loads of spectra with no SNNAME and just _$MJD.spec in ./spectra/

[RobFirth]

Running ./specphase generates loads of spectra with no SNNAME and just _$MJD.spec in ./spectra/

[sprajs]

This is the most bizarre bug I've dealt with in ages, there is nothing at the point where the bug occurs that could cause it to push an empty data structure into an unordered_map<>. The more interesting thing is that the empty structure is added when the first spectrum for SN1998bw is processed regardless of what order the spectra are read in.

[RobFirth]

Can confirm that 1998bw seems to be the issue with me too - specfit segfaults on data/spec/SN1998bw/1998bw_-10.61.txt, which I assume has something to do with it? Is that what you mean by 'first' or do you mean the first in memory?

[RobFirth]

and is it just me or does SN1998bw never get it's name prepended to the spectra?

[sprajs]

By first I mean first in memory, unordered_map<> is a hash table so the order in which the elements are stored is always random. I get a segfault in specfit too. I wonder if it's something to do with the format for that specific file only, I will investigate.