Pandas warnings fixes #1089
Pandas warnings fixes #1089
Conversation
|
My IDE complains if the docstring is not correct. I know, this should be a different PR. Please forgive my laziness 😄 |
|
Somehow the values for the OffsetConverter are build differently so one test fails with my changes. I could not find the reason. |
| @@ -43,7 +54,7 @@ def sequence(iterable_or_scalar): | |||
| if isinstance(iterable_or_scalar, abc.Iterable) and not isinstance( | |||
| iterable_or_scalar, str | |||
| ): | |||
| return iterable_or_scalar | |||
| return np.array(iterable_or_scalar) | |||
There was a problem hiding this comment.
There is a problem with this, as sequence is used to create sequences of practically infinite length from scalar values. If you return a numpy.array instead, the following will happen:
# current implementation
s = sequence(sequence([5]))
s[0] = 5
s[5] = 5
# your suggestion
s = sequence(sequence([5]))
s[0] = 5
s[5] -> out of range
Note thatthis is just a workaround. chaned casts chould either work or be prohibited.
| slope = sequence(slope[0]) | ||
| offset = sequence(offset[0]) | ||
| slope = slope[0] | ||
| offset = offset[0] | ||
|
|
There was a problem hiding this comment.
Thank you for jumping in here @p-snft. I was actually wrapping my head around this sequence stuff the first time I saw it when implementing the changes for the OffsetConverter and it took me quite a while to figure out what was going on. I see the practical point of the sequences, but I was wondering, if we could not just use scalars or numpy arrays instead in general. Because if you (by accident) provide a half length sequence (with respect to the timeindex) you would end up with some kind of nonsense anyway, right? So why not let people either pass scalars or correct-length lists or arrays?
There was a problem hiding this comment.
The problem is that we do not know the correct length before the Entity is added to the EnergySystem. An option would be to just let scalar stay scalars until we need to have arrays.
A lot of inputs in solph can be single values or sequences.
Recently a
FutureWarningfrom pandas warns that Series will not work in the future. A small change in our plumbing module will keep the old behaviour and avoid theFutureWarning