#2501 CDF editor dialog implies you can select a set of parameters to plot, independently

[Jeremy Faden]

The CDF GUI implies you can select a set of parameters to plot, independently. Load https://cdaweb.gsfc.nasa.gov/sp_phys/data/mms/mms1/fpi/fast/l2/des-moms/2017/06/mms1_fpi_fast_l2_des-moms_20170617080000_v3.3.0.cdf and then in the GUI select four or five parameters. It will crash with an IndexOutOfBoundsException, not a warning that only one parameter can be plotted at a time.

[Jeremy Faden]

See rte_1011208926_20230115_181507*.xml .

[Jeremy Faden]

This came up again when an RTE was submitted where a configuration was specified that couldn't be performed. Connor at Iowa was just using this feature last week, and I realized how much interest there is in the feature.

[Jeremy Faden]

rbsp-b_wfr-spectral-matrix_emfisis-l2_20170101_v1.6.4.cdf?EwEw&x=EwEw results in "java.lang.IllegalArgumentException: DEPEND_0 should have only one index or must be a bins ([n,2]) dataset."

[Jeremy Faden]

Some rules that should be implemented:

• when CDF variable is rank 1, only rank 1 variables can be picked for X, or rank 2 n,2 "bins" dataset
• length() of variables of X must match length of CDF variable
• length() of variables of Y must match length of CDF variable, or length(0) of rank 2 CDF variable.

[Jeremy Faden]

The "sort alpha" checkbox only sorts the "select variable" tab, but not the X and Y tabs.

[Jeremy Faden]

Note you can select multiple "dependent" parameters, Y in Y(Time), which was added to support HAPI servers.

[Jeremy Faden]

I did a bunch of cleaning up the gui, for example adding completions to X= and Y=, and loading the correct metadata when X= is used.

[Jeremy Faden]

Also there was a bug where the info about the variable wouldn't update when you clicked on new values. Also multiple parameters (L;MLT) added for HAPI wouldn't show in the dialog.

[Jeremy Faden]

Note, this is all going to have to happen to the NetCDF dialog too.

[Jeremy Faden]

The submission rte_2101508540 shows where the last production release didn't have a branch for non-virtual variables when BX_GSE;BY_GSM;BZ_GSM is selected. This was fixed several days ago. Thanks Dante!