Make CSV reader automatically use column 1 as default X Axis rather than using index for X Axis

ParaView Support
1

I have a finite element program that writes out the time-histories of user-defined rigid wall forces to a CSV file. The first line of the CSV file contains the column labels, and for a file with 7 rigid wall definitions, it looks like this:

(note this is all one line … each entry takes up 19 spaces, separated by a comma)

         "TIME",        "RW_1_FMAG",        "RW_2_FMAG",        "RW_3_FMAG",        "RW_4_FMAG",        "RW_5_FMAG",        "RW_6_FMAG",        "RW_7_FMAG",           "RW1_FX",           "RW1_FY",           "RW1_FZ",           "RW2_FX",           "RW2_FY",           "RW2_FZ",           "RW3_FX",           "RW3_FY",           "RW3_FZ",           "RW4_FX",           "RW4_FY",           "RW4_FZ",           "RW5_FX",           "RW5_FY",           "RW5_FZ",           "RW6_FX",           "RW6_FY",           "RW6_FZ",           "RW7_FX",           "RW7_FY",           "RW7_FZ"

Each time stepstep is then output on a single line looking like this (line 2):

0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000, 0.0000000000E+000

(Note, the above is just an ES19.10E3 format from Fortran, with a comma separating the entries)

The line 2 data repeats until the end of the file so it’s just a plain and simple CSV file.

To visualize my data in ParaView, I open the file, choose “CSV Reader”, and then open it in a Line Chart View.

The annoying part is then that every time, I have to deselect “Use Index for XAxis” and then select the TIME data for the “X Array Name”, which is frustrating because I know I’ll typically want to plot it against the TIME data in column 1.

More importantly, this is frustrating because I have a similar CSV file containing nodal time-history data that behaves as I’d like it to. Again, TIME is in column 1, and then accelerations, displacements, etc. are in the remaining columns, with the same format as above. But with that file it automatically uses column 1 as the default XAxis.

I have played around with the rigidwall data file, and it appears if I name the columns slightly different, then this can lead to it using TIME as the XAxis, but I can’t figure out what rules the CSV Reader is applying to determine when to do this and when not to.

Can someone please clarify the guidelines for how to make the CSV Reader consistently interpret the data so the user-experience of opening the file is the same every time?

I’m happy to explain more as needed. Thanks.

P.S., I’m running ParaView 5.10.1 on Windows 10.

2

HI @jman87

Indeed, certain names are automatically detected as “valid” X array names:
in_extents;Time;time;arc_length;XArray;x_array;Frequency

You can configure this in Edit → Settings → Represented Attributes → Chart Default X Axis

Best,

3

Wow! :rofl: So it’s literally as simple as changing my header from “TIME” to “Time”!!!

Perhaps Kitware might consider adding “TIME” to that list for all of us that like to have our legends yell at us. :grin:

Thanks @mwestphal

4

Since the list can now be configured, we would rather let user create their own configuration.