### Excitation function from file or interpolated function

Posted:

**Thu 10 Sep 2020, 13:35**Hello Thorsten,

a functionality that reads arbitrary excitation waveforms from a file would be nice / advantageous, because sometimes it is almost impossible to come up with approximations that could be described as polynomials (which would be converted to the string in Octave), if for example real, filtered measurements shall be used to feed the algorithm.

--> from the values (time series) in the file one could for example generate a interpolated function (like scipy.interpolate.interp1d in Python),

which would serve as a "analytical" source function that the FDTD algorithm could evaluate in every update step ...

I wonder, if this would be a lot of work to implement and where would be the place to add this in the source code?

Reason: I had to implement source currents J, which I achieved by implementing the curling ∇xH, hence four magnetic sources corresponding to ∇xH=J.

Then I need many hundred of such sources stacked, that together describe the return stroke of a lightning channel. This works for analytical currents, but if real measurements want to be inserted (which are for instance difficult to describe by analytical approximations) this wouldn't be possible.

Thanks in advance!

Best regards,

Hannes

a functionality that reads arbitrary excitation waveforms from a file would be nice / advantageous, because sometimes it is almost impossible to come up with approximations that could be described as polynomials (which would be converted to the string in Octave), if for example real, filtered measurements shall be used to feed the algorithm.

--> from the values (time series) in the file one could for example generate a interpolated function (like scipy.interpolate.interp1d in Python),

which would serve as a "analytical" source function that the FDTD algorithm could evaluate in every update step ...

I wonder, if this would be a lot of work to implement and where would be the place to add this in the source code?

Reason: I had to implement source currents J, which I achieved by implementing the curling ∇xH, hence four magnetic sources corresponding to ∇xH=J.

Then I need many hundred of such sources stacked, that together describe the return stroke of a lightning channel. This works for analytical currents, but if real measurements want to be inserted (which are for instance difficult to describe by analytical approximations) this wouldn't be possible.

Thanks in advance!

Best regards,

Hannes