Conversation
adding an option to mathics/test.py for storing errors in a log file. Fixing some tests. Fixing -Infinity. adding an option to mathics/test.py for storing errors in a log file. Fixing some tests. Fixing -Infinity.
mmatera
force-pushed
the
fixtimesinfinity
branch
from
86c1979 to
d46fdb1
Compare
| "DirectedInfinity[a_?NumericQ] /; N[Abs[a]] != 1": "DirectedInfinity[a / Abs[a]]", | ||
| "DirectedInfinity[a_] * DirectedInfinity[b_]": "DirectedInfinity[a*b]", | ||
| "DirectedInfinity[] * DirectedInfinity[args___]": "DirectedInfinity[]", | ||
| "DirectedInfinity[0]": "DirectedInfinity[]", |
There was a problem hiding this comment.
flake8 correctly informs me that "DirectedInfinity[0]" appears twice. See below in line 1116. The second occurrence (that is, not this one), will be the one used.
If that is what is desired then it would be better to remove this line.
rocky
left a comment
rocky
left a comment
There was a problem hiding this comment.
LGTM - but please note the comment about the duplicate dictionary key that appears.
|
Thanks! I fixed this here. Shall I merge this if this pass the tests? |
Yes - and then if we could rebase (not merge) master onto the extensecmptests branch that would be good. When we have several branches that all want to go into master with the exact code, it is hard to understand what is going on individually because of the duplication which obscures the essential part of the PR. Thanks. |
|
Indeed. I am trying to keep the extensecmptests updated with the changes in fixtimesinfinity, and make that more tests pass. What I wasn't sure is if it is better to merge the parts that are working now, or to have a merge with the complete fix. |
|
Ok - then let's merge this. I'll do the rebase after the merge. |
2 participants
This PR tries to fix errors and inconsistencies in expressions involving products of complex numbers and Infinity quantities.