[BUG] Time reversal

[djps]

Describe the bug
Time reversal with cartesian data does not work: there are two possible issues:

• the output of simulations is in a different order to matlab.
• interp_cart_data may not be correct

To Reproduce
Steps to reproduce the behavior:
run
https://gist.github.com/djps/1f9ea4bc080b601a45c4e81f2ed7015a
and compare the behaviour to matlab: there are three simulations which are run. the first computes the data which is time reversed. The second is time reversal, the third is time reversal with interpolation.

python on left - matlab on the right.

Expected behavior
The output from the first simulations does not seem to match: it is certainly in the wrong order. The seond simulation is able to perform time reversal. The third simulation is incorrect as the interpolated data from interp_cart_data is wrong: this could be because the sensor data is in the wrong order or the function itself is incorrect and testing for the interp_cart_data may not be effective.

Desktop (please complete the following information):

• OS: Windows 11

Additional context

[waltsims]

For me the second simulation (time-reversal) in your gist does not run on master. Could there be changes in your local version of k-wave-python that I am missing?

[djps]

@waltsims - are some differences in the time reversal class and the matlab code. For validation, there should some consistency with:

• https://github.com/ucl-bug/k-wave/blob/aef4ce6852807a33a8b0b4507cb643d8987bb73e/k-Wave/kspaceFirstOrder2D.m#L669
• https://github.com/ucl-bug/k-wave/blob/aef4ce6852807a33a8b0b4507cb643d8987bb73e/k-Wave/kspaceFirstOrder2D.m#L1045

so that the final output computed up to the penultimate time, but the has sensor data overlaid.