PythonQuery without Addplots()

[chrspy1]

I am using the python visit cli, and I would like to pull data from specific arrays, do some math, and output an answer that is a single value.

I understand that I can use a python Query to do this. For example, I can do something like:
PythonQuery(file='q.py',vars={"var1","var2"})

However, such a query requires "an active and nonhidden plot" to work. For my application, I only care for a single value (for example sum(var1*var2). From this, I have two questions.

1. How can I use the VTK methods in Visit's python scripting to pull values directly into arrays, and perform my operations using numpy? For example, how would I do something like:

      import visit
      import numpy as np 
      # add no win argument to run because I dont care about plots
      AddArgument("-nowin")
      Launch()
      OpenDatabase("foo.exo")
      .. some stuff to create "ds" so I can use GetPointData() to pull data without having to plot ... 
      v1 =np.numpy(ds.GetPointData().GetArray("my_v1"))
      v2 =np.numpy(ds.GetPointData().GetArray("my_v2"))
      output = np.sum(v1*v2)
      print("the answer is %E" %(output))

2. Is there a workaround to running PythonQuery without needing to plot?
   Thanks,
   Chris

[JustinPrivitera]

VisIt doesn't read the data from files until you ask it to plot. Before that, it just reads root file metadata. So asking it to perform operations on data in the files before you have plotted anything doesn't work.

There is good information here on python queries: https://visit-sphinx-github-user-manual.readthedocs.io/en/develop/using_visit/Quantitative/Query.html#python-queries

There is a code example there that does cell_data = ds_in.GetCellData().GetArray(self.input_var_names[0]), which is similar to what you were asking for with v1 =np.numpy(ds.GetPointData().GetArray("my_v1")).