search

USRA-STI / gdt-fermi

Gamma-ray Data Tools - Fermi mission components
Apache License 2.0
3 stars 7 forks source link

Extension missing in re-written files #37

Closed nyctophile-aayu closed 3 months ago

nyctophile-aayu commented 3 months ago

Hi, I have tried the following notebook to generate to process following steps:

  1. Load the tte.fit file
  2. Process time binning
  3. Then process energy binning using (combine_by_factor) method
  4. Re-write the file.

I have tried re-write into .fit, .pha2 files both.

But the issue is: in the original .fit file, in the events extension, Time & Pha columns are listed, which should be there in the re-written .fit file. But in the re-written tte.fit file there is no Events extension present, instead in the Spectrum extension, Channel,Counts,Quality columns are present. This updated file when processed for the spectral study, using:

n0 = Cspec.open('data1/glg_tte_n0_bn220627890_v00.fit')

generating error: KeyError: "Key 'ENDTIME' does not exist."

Also, in the previously converted .pha2 file from GBM tools used to process for spectral analysis using:

n0 = Cspec.open('data1/glg_tte_n0_bn220627890_v00.pha2')

works fine. But that file is not energy binned. But if I process the .fit file for both time and energy binning and the re-write into .pha2 file then also inspite of having columns: Counts, Exposure, Quality, Time, Endtime, in the Spectrum extension, again Channel,Counts,Quality columns are present in the Spectrum extension. This updated file when again process for spectral study then it is causing the same error.

Is there a way to obtain these columns: Counts, Exposure, Quality, Time, Endtime, in the Spectrum extension in the updated .pha2 or .fit file? If yes, please suggest how?

I am attaching the files for your reference. pha_files.zip

AdamGoldstein-USRA commented 3 months ago

Hi @nyctophile-aayu:

When you bin TTE in time using the to_phaii() method, it creates a Phaii object (also called Cspec or Ctime for the two different types of GBM Phaii objects). This is a fundamentally different datatype than TTE. TTE is an event list, where the time and channel number for each count is recorded. When you bin in time, it is no longer an event list. It is a time series of spectra. That is what a PHA2 file is (and what a Phaii object represents).

When you integrate a TTE or Phaii object over time, you get a Pha object. What distinguishes a Pha object from Phaii is that Pha is only a single spectrum (because you integrated over time), while Phaii contains a time series of spectra.

The first problem you mention is due to the fact that you binned TTE in time and created a PHA2 file, therefore it won't have the individual count information. The second problem you mention is because you integrated the PHA2 over time to produce a PHA file, which will have different columns than a PHA2 file.

I encourage you to look at some of the documentation to clear up some of these issues. PHA documentation PHAII documentation TTE documentation