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
Preferably these node tables are kept None, such that it is clear that they are not present, e.g. in _nodes() from #1270. With #1279 we now work around this by not only checking node.df is None but additionally attr.node.df.empty.
In #1276 I came across this issue and try to adapt the filter function, which caused several other issues.
With the workaround the issue itself is not so pressing, though it is worth seeing if we fully understand and perhaps can simplify some of the validation logic causing the unwanted side effects seen in #1276.
The text was updated successfully, but these errors were encountered:
This supersedes #1276 by working around #1278 and still enabling the
round tripping tests.
Also simplifies `__assert_equal` a bit and stimulates
`Model.read(toml_path)` usage over `Model(filepath=toml_path)`.
Preferably these node tables are kept None, such that it is clear that they are not present, e.g. in
_nodes()
from #1270. With #1279 we now work around this by not only checkingnode.df is None
but additionallyattr.node.df.empty
.In #1276 I came across this issue and try to adapt the
filter
function, which caused several other issues.With the workaround the issue itself is not so pressing, though it is worth seeing if we fully understand and perhaps can simplify some of the validation logic causing the unwanted side effects seen in #1276.
The text was updated successfully, but these errors were encountered: