Menu
▾
▴
#2467 Old LANL CDF kludge is messing up PSP data
There's an old LANL CDF kludge, where if a DEPEND_1 has two records and can be interpreted as min and max, then the average is returned. This is causing problems when reading some Parker Solar Probe data, which is a CDF file with just two records.
Discussion
I've put in a check to see that this would result in a valid dataset, checking that the number of records expected isn't two, and this allows the code to do more.
The PSP data was https://cdaweb.gsfc.nasa.gov/sp_phys/data/psp/fields/l2/dfb_dbm_vac/2021/psp_fld_l2_dfb_dbm_vac_2021011812_v02.cdf?psp_fld_l2_dfb_dbm_vac1
Also, QDataSet has been able to represent min,max ranges for a while, and this kludge might be updated to support this.
Another issue that's related: Sadie and Cindy were using the "X" and "Y" selections of the CDF editor, without selecting a variable for the first dimension (Z). This should be allowed, since it's explicit. And, the list of parameter names was not the same on the three lists, and I need to remind myself why this is.
And in the waveform case, the three datasets should be flattened to allow more easy use, and since DEPEND_0 cannot be rank 2 and a BUNDLE is of rank 1 datasets.