Marco Berzborn
Marco Berzborn
I'm honestly a bit confused what the purpose of this function should be if it's a random Poisson distributed sequence, so it's a kind of noise process with random temporal...
I would propose another idea: We could simply add a physical unit property to the class objects. If none is given, the current behavior is kept unchanged. If a unit...
I don't see the problem, could you please elaborate?
> I thought about that as well. It could be done in the style of solutions 1-3. But maybe the unit itself would not be enough. For example the unit...
Agreed ;) My primary concern would be to avoid unintuitive/inconsistent/hacky ways of doing things within pyfar, such as having to specify `dB=False` to get a correct logarithmic plot in dB...
The outcome of the discussion in the last meeting was to avoid writing dB quantities to `FrequencyData` or `TimeData` to avoid erroneous behavior within pyfar. In cases where dB quantities...
I think Simon is referring to libraries such as pydantic.
There's a separate issue for the overlapping tick labels https://github.com/pyfar/pyfar/issues/451
1. The addition of the `LTISystem` indeed looks like a good addition. I have to admit that my expertise on state space systems is very limited, so I won't comment...
I guess if it's also useful to quantify noise contamination in arbitrary signals it might fit the base package well. If it's more specific we could dicuss other options (for...