You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In #168, it is discussed that the the strict bounds should be relaxed.
But it is important to add a verification that the photosphere bounds defined by the user must match the values available in the input atmosphere tables. Otherwise, a clear error message should be displayed.
Similarly, there should be a check that the dimensions of the photosphere match those of the input atmosphere tables.
The text was updated successfully, but these errors were encountered:
sguillot
changed the title
Add a check that the photosphere bounds match input atmosphere
Add check that the photosphere bounds and dimensions match the input atmosphere
Oct 14, 2022
I have thought a bit about how to do this in practice, and I see now 3 different options:
We can add this type of check in the installable part of X-PSI (but outside the atmosphere extension), but it may require abandoning some of the flexibility with different atmosphere extensions that X-PSI currently has. We would e.g. have to enforce which of the atmosphere table parameters is temperature (first one in the table?) and that it must be in the format of "logT" as currently with nsx..... However, probably this check could be such, that it can be overwritten by the user.
This check could be done in the atmosphere extension itself (i.e. hot.pyx), then it would indeed be specific to the used extension. In that case, need to pass the information about the hard bounds to the functions in that file.
This check could also be done in the scripts that we use to run X-PSI (easiest option).
Another question is whether this needs to be an error rather than a warning. One might also want to keep the possibility of extrapolating outside of the atmosphere table bounds, or is that always a very bad idea?
Also, the atmosphere table bounds can be multi-dimensional in terms of X-PSI parameters, e.g. surface gravity depending both on mass and radius, or a luminosity parameter depending on temperature, mass, and radius. Then looking at the hard bounds of a single X-PSI parameter is not very useful.
In #168, it is discussed that the the strict bounds should be relaxed.
But it is important to add a verification that the photosphere bounds defined by the user must match the values available in the input atmosphere tables. Otherwise, a clear error message should be displayed.
Similarly, there should be a check that the dimensions of the photosphere match those of the input atmosphere tables.
The text was updated successfully, but these errors were encountered: