# Exporting Data from Descriptive Statistics

**URL:** https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811
**Category:** ParaView Support
**Created:** [November 19, 2022, 4:30pm UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811 "2022-11-19T16:30:41Z")
**Posts on this page:** 5
**Page:** 1

<div class="post-metadata">

### Author: ![jgd2203](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/j/46a35a/32.png) [@jgd2203](https://discourse.paraview.org/u/jgd2203)
#### Post date: [November 19, 2022, 4:30pm UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811/1 "2022-11-19T16:30:41Z")

</div>

I have used the tracing function to create a script to extract some statistics (using descriptive statistics filter) based on the surfaces of my mesh. When I play back in Windows it seems to work fine.  
Basic Steps:  
(1) Read Tecplot Binary  
(2) Extract Block to get desired surfaces  
(3) Apply Descriptive FIlter  
(4) Save Data

when I move the script over to our cluster (RHEL8) the script executes without error and creates the csv file (using SaveData, FieldAssociation=‘Row Data’) but the file is empty.

I recreated the trace directly on the RHEL head node interactively and it still writes a empty file. If I execute it using pvbatch or a python shell in the PV GUI - it runs w/o error but creates an empty file.

What am I missing here? Thanks!

Tested with PV 5.10 and 5.11 and each behave the same way.

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 21, 2022, 7:50am UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811/2 "2022-11-21T07:50:15Z")

</div>

HI @jgd2203

Is the path to the tecplot file correct on the cluster ?

Best,

---

<div class="post-metadata">

### Author: ![jgd2203](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/j/46a35a/32.png) [@jgd2203](https://discourse.paraview.org/u/jgd2203)
#### Post date: [November 21, 2022, 2:05pm UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811/3 "2022-11-21T14:05:32Z")

</div>

yes, I verified via print statement and ls cmd and as I said I did a brand new trace (on the cluster) w/o editing anything and on playback it behaved the same way. Regardless, I found a work around. I just save all the data before using the filter and compute statistics in python. I suppose it would still be good to figure this out as it seems like a useful filter.

On a related but different topic - how do I get the zone names to a pyhton list? I am using a Tecplot binary reader and it seems to read the datatype as vtkMultiBlock. I can of course see them after the extractBlock filter in the GUI information tab - but not sure how to get these into a list.

---

<div class="post-metadata">

### Author: ![jgd2203](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/j/46a35a/32.png) [@jgd2203](https://discourse.paraview.org/u/jgd2203)
#### Post date: [November 29, 2022, 4:04pm UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811/4 "2022-11-29T16:04:37Z")

</div>

Just following up as there has been no resposnse to the above.

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 29, 2022, 4:15pm UTC](https://discourse.paraview.org/t/exporting-data-from-descriptive-statistics/10811/5 "2022-11-29T16:15:31Z")

</div>

`GetDataInformation().DataInformation.GetBlockName(idx)` should work.
