[hz283]

Dear All,

I use paraview to calculate the temporal statistics - mean and RMS. I have 500 time instants and of course I will read them all into paraview. However, sometimes I would like to calculate the statistics from 100th to 500th time instants. How can I choose the starting and ending time instants in this filter? Thank you very much if someone can give me hints. Thanks again.

h

[wyldckat]

Hi hz283,

AFAIK, ParaView is a bit limited when it comes to manipulating time. Not even ParaView 4.0.1 has got that feature, from what I can see.

They've focused on manipulating time from the "animation" point of view, not as much from the temporal data point of view.

The only manipulation that I'm aware that is 100% bulletproof is to:

1. Export the data to VTK files, where each time instance is stored in a single VTK file.
2. Then remove the files that relate to the temporal times you don't want to use.
3. Then open in ParaView the resulting group of files, e.g. "data_files..vtk".

Best regards,
Bruno

[openfoammaofnepo]

Thank you. Hope Paraview can have this feature in the future, which will make life more happy.

Originally Posted by wyldckat

Hi hz283,

AFAIK, ParaView is a bit limited when it comes to manipulating time. Not even ParaView 4.0.1 has got that feature, from what I can see.

They've focused on manipulating time from the "animation" point of view, not as much from the temporal data point of view.

The only manipulation that I'm aware that is 100% bulletproof is to:

1. Export the data to VTK files, where each time instance is stored in a single VTK file.
2. Then remove the files that relate to the temporal times you don't want to use.
3. Then open in ParaView the resulting group of files, e.g. "data_files..vtk".

Best regards,
Bruno

[flowAlways]

Hello,
I was looking for something similar but removing the files that I don't need in the computation of the statistics was not so convenient
(although could be automated by writing a simple shell script for large dataset).
An alternative that I use is by specifying the files to be considered for averaging in the ensight gold format.

For example in the format below by changing "number of steps" to 1000, "filename start number" to 4001 and removing the first 4000 time
steps in the "time values" list allows to load only the last 1000 time steps and then average of the last 1000 time steps(only) can be computed
without actually removing the other time files from the directory.
Note: only loading this file will instruct the paraview to load the specified dataset.

FORMAT
type: ensight gold

GEOMETRY
model: pF.geo
VARIABLE
vector per node: 1  velocity pF.vel****
scalar per node: 1 vorticity pF.vor****

TIME
time set:              1
number of steps:       5001
filename start number: 1
filename increment:    1
time values:
    0
    0.01
    . 
    .
    50

[eternusvia]

I found the best way to get time-averaged statistics for a subset of time is to use the "Extract Time Steps" filter.

Filters > Temporal > Extract Time Steps

Then you add the range of time steps you want to extract.