Strange Errors for Post-Processing Roving Hammer Data

[Nidish96]

Firstly, kudos to the developers - this is a really well designed package for most things modal, thanks a lot!

I've encountered a very strange bug while doing polyreference fit on a roving hammer dataset. I had a plate specimen with 4 accelerometers attached to the corners and roving hammer data for a 4x4 grid of (16) points.

Calling sdynpy.PolyPy_GUI on this FRF works but whenever it tried to synthesize it throws the following error:

Traceback (most recent call last):
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/modal/sdynpy_polypy.py", line 801, in compute_shapes
    self.shapes, self.resynthesized_frfs, self.frfs_synth_residual, self.frfs_synth_kernel = (
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<8 lines>...
            None if self.all_frequency_lines_checkbox.isChecked() else self.lines_at_resonance_spinbox.value(),
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/modal/sdynpy_modeshape.py", line 509, in compute_shapes_multireference
    frfs_resynthesized = (output_shape.compute_frf(
                          ~~~~~~~~~~~~~~~~~~~~~~~~^
        frequencies,
        ^^^^^^^^^^^^
        responses=experimental_frf.response_coordinate[:, 0],
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        references=experimental_frf.reference_coordinate[0, :],
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        displacement_derivative=displacement_derivative)
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/core/sdynpy_shape.py", line 516, in compute_frf
    reference_shape_matrix = flat_self[references]  # nm x ni
                             ~~~~~~~~~^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/core/sdynpy_shape.py", line 194, in __getitem__
    raise ValueError(
        'Not all indices in requested coordinate array exist in the shape\n{:}'.format(str(extra_keys)))
ValueError: Not all indices in requested coordinate array exist in the shape
['2Y+' '3Y+' '5Y+' '6Y+' '7Y+' '8Y+' '9Y+' '10Y+' '11Y+' '12Y+' '14Y+'
 '15Y+']

Scale for mode 1: (-3.715743983207786-3.3003698788755154j)
Scale for mode 2: (4.139374910612414+4.737157125173919j)
Traceback (most recent call last):
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/modal/sdynpy_polypy.py", line 801, in compute_shapes
    self.shapes, self.resynthesized_frfs, self.frfs_synth_residual, self.frfs_synth_kernel = (
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<8 lines>...
            None if self.all_frequency_lines_checkbox.isChecked() else self.lines_at_resonance_spinbox.value(),
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/modal/sdynpy_modeshape.py", line 509, in compute_shapes_multireference
    frfs_resynthesized = (output_shape.compute_frf(
                          ~~~~~~~~~~~~~~~~~~~~~~~~^
        frequencies,
        ^^^^^^^^^^^^
        responses=experimental_frf.response_coordinate[:, 0],
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        references=experimental_frf.reference_coordinate[0, :],
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        displacement_derivative=displacement_derivative)
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/core/sdynpy_shape.py", line 516, in compute_frf
    reference_shape_matrix = flat_self[references]  # nm x ni
                             ~~~~~~~~~^^^^^^^^^^^^
  File "/home/nidish/.local/lib/python3.13/site-packages/sdynpy/core/sdynpy_shape.py", line 194, in __getitem__
    raise ValueError(
        'Not all indices in requested coordinate array exist in the shape\n{:}'.format(str(extra_keys)))
ValueError: Not all indices in requested coordinate array exist in the shape
['2Y+' '3Y+' '5Y+' '6Y+' '7Y+' '8Y+' '9Y+' '10Y+' '11Y+' '12Y+' '14Y+'
 '15Y+']

It basically says that all the indices except for the indices corresponding to my sensor locations can't be found. I tried to look at the code a little bit but I couldn't follow much unfortunately.

A Workaround

After snooping around in the code I realized that it kind of expects the response_dofs to be a superset of the reference_dofs . So in my code I just swapped the reference and responses (this is okay I guess since the linear FRF is symmetric anyways). Now everything works!

This is okay for me, but I certainly think that this must be a relatively simple bug.

Here's how I'm reading in the data for the workaround.

dat = sdn.uff.readuff('./Frequency Responses.uff')

# BUG WORKAROUND: Reversing the references and responses
for d in dat[58]:
    rdir = d.reference_direction
    rnd = d.reference_node

    d.reference_direction = d.response_direction
    d.reference_node = d.response_node

    d.response_direction = rdir
    d.response_node = rnd

Files to reproduce the bug

I've uploaded the dataset (as a UFF 58 file), a text file with node positions, and a python script using SDynPy here. This bug can be reproduced by commenting out lines 51-59 and running the script.
https://drive.google.com/drive/folders/1kTd67nOCJuEFnDVT8-OBlKnzplRy6f0R?usp=sharing