# Using Python to load multiple CSVs

**URL:** https://discourse.paraview.org/t/using-python-to-load-multiple-csvs/12105
**Category:** ParaView Support
**Tags:** python
**Created:** [May 16, 2023, 4:45am UTC](https://discourse.paraview.org/t/using-python-to-load-multiple-csvs/12105 "2023-05-16T04:45:38Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![arkottke](https://discourse.paraview.org/user_avatar/discourse.paraview.org/arkottke/32/11617_2.png) [@arkottke](https://discourse.paraview.org/u/arkottke)
#### Post date: [May 16, 2023, 4:45am UTC](https://discourse.paraview.org/t/using-python-to-load-multiple-csvs/12105/1 "2023-05-16T04:45:38Z")

</div>

I am trying to plot fault planes and epicenters. I was thinking that this would be a nice exercise to learn paraview. My basic process is to create CSV files of the fault coordinates. I then wrote paraview script to take all of these and create Delaunay2D surfaces of the faults. Here is the script:

```python
from pathlib import Path
from paraview.simple import *

root = Path( __file__ ).parent

table = CSVReader(FileName=str(root / 'data/events.csv'), guiName='Events')
pts = TableToPoints(table, XColumn='easting', YColumn='northing', ZColumn='depth')
Show()

dp = GetDisplayProperties(pts)
dp.Representation = 'Point Gaussian'

for fpath in (root / 'exported').glob('source-*.csv'):
    table = CSVReader(FileName=str(fpath), guiName=fpath.stem)
    pts = TableToPoints(table, XColumn='easting', YColumn='northing', ZColumn='depth')

    surf = Delaunay2D(pts)
    surf.ProjectionPlaneMode = 'Best-Fitting Plane'

    Show()

renderView = GetActiveView()
ag = renderView.AxesGrid
ag.Visibility = 1

ag.XTitle = 'Easting (km)'
ag.YTitle = 'Northing (km)'
ag.ZTitle = 'Depth (km)'

ResetCamera()
Render()

```

And the resulting output:

 ![image](https://discourse.paraview.org/uploads/default/original/2X/c/c55683719d96736946e6a4bbfdaca4e30eb364e3.jpeg)

I have a few questions/comments:

1. Is there a better way to create the faults? At the surface, the triangulation fills in areas when the faults curve. I was thinking I might be able to put all of the faults into one file.
2. The current approach creates a pretty complicated Pipeline. Is there a good way to create a tree view or group sources and filters? I see this is done when I do things manually, but not programmatically.
3. I can adjust the size epicenters (spheres) by “Use Scale Array” and selecting “magnitude”, but I couldn’t figure out how to do this via the Python interface. How do people usually figure how the relationship between the buttons and the Python?

---

<div class="post-metadata">

### Author: ![jourdain](https://discourse.paraview.org/user_avatar/discourse.paraview.org/jourdain/32/11_2.png) [@jourdain](https://discourse.paraview.org/u/jourdain)
#### Post date: [May 16, 2023, 2:29pm UTC](https://discourse.paraview.org/t/using-python-to-load-multiple-csvs/12105/2 "2023-05-16T14:29:00Z")

</div>

You should use [the trace to convert UI interaction into code](https://discourse.paraview.org/t/how-to-figure-out-a-python-code-using-the-python-trace-mechanism/4186).