The following Science Tools are tried out on a Linux machine: gtselect, count_map, exposure_cube, and exposure_map. A dataset was obtained through the GSSC web page for data retrieval for this checkout.
Summarized below are my findings.
1) Not conforming the parameter interface standard proposed by James Peachey.
gtselect
- input_file should be renamed to evfile.
- evtable missing.
- output_file should be renamed to outfile.
count_map
- infile should be renamed to evfile.
- table should be renamed to evtable.
- clobber and mode should default to h (hidden).
exposure_cube
- infile should be renamed to scfile.
- table should be renamed to sctable.
- clobber and mode should default to h (hidden).
exposure_map
- infile should be renamed to scfile.
- table should be renamed to sctable.
2) The version number is not displayed by count_map, exposure_cube, and exposure_map. They report like "count_map: INFO: This is count_map version unknown" (for the case of count_map) and no specific version number is displayed for users.
3) Poor explanations of parameters for exposure_map. The meaning of "layers", "emin", and "eratio" are not clear. Explain them somewhere, e.g., in their prompts, on the web page(s), or in its user guide.
4) A concern: exposure_cube takes a long time to complete its computation; it took 790 s for one-day data on a Linux machine at GSSC (called glitch). If it grows linearly with the length of observation, it is estimated to take 3.3 days for one-year data.
5) Many issues raised in Checkout 1 are not resolved yet.