Bidirectional gridPathCells #1111
Conversation
|
|
256606e to
d3b17f1
Compare
d3b17f1 to
8bb9664
Compare
isaacbrodsky
left a comment
There was a problem hiding this comment.
When I look at the path generated by this call, I am noticing a cell that does not seem like it should be in the output:
8411b61ffffffff
8411b6bffffffff
8411b47ffffffff
8411b43ffffffff
8411b5dffffffff
8411b59ffffffff
841f49bffffffff
8411a6dffffffff
8411a69ffffffff
8411a45ffffffff
8411a41ffffffff
8411a4bffffffff
841f693ffffffff
841f69bffffffff
841116dffffffff
8411169ffffffff
8400db7ffffffff
841114dffffffff
8411149ffffffff
8411335ffffffff
8411331ffffffff
841133bffffffff
8411317ffffffff
8411313ffffffff
841104dffffffff
8411049ffffffff
8411223ffffffff
841123dffffffff
8411239ffffffff
8411215ffffffff
8411211ffffffff
841121bffffffff
8408d27ffffffff
8408d23ffffffff
8408d3dffffffff
8408d39ffffffff
8408d03ffffffff
8408d1dffffffff
8408d19ffffffff
8408c25ffffffff
8408c21ffffffff
8408c2bffffffff
8408c07ffffffff
8408c03ffffffff
8408c1dffffffff
8408c19ffffffff
8408c53ffffffff
8408eadffffffff
8408ea9ffffffff
8408e85ffffffff
8408e81ffffffff
8408e8bffffffff
8408ed7ffffffff
8408ed3ffffffff
84094b5ffffffff
84094bdffffffff
84094abffffffff
8409485ffffffff
840948dffffffff
84094c7ffffffff
84094c1ffffffff
84094c9ffffffff
84096a3ffffffff
84096abffffffff
8409685ffffffff
840968dffffffff
84096ebffffffff
84096c5ffffffff
84096cdffffffff
840164dffffffff
8401641ffffffff
8401643ffffffff
8401655ffffffff
8401657ffffffff
8401619ffffffff
840161bffffffff
8407365ffffffff
84016c9ffffffff
84016cbffffffff
84016ddffffffff
84016d1ffffffff
84016d3ffffffff
The cell in question is 8400db7ffffffff
That said, this issue does not seem to be introduced in this PR. I noticed that when I tried the reverse, I got a different apparently defective cell: 84009b3ffffffff
|
@isaacbrodsky For |
Sorry, my mistake. I meant to check that this solves for a case where it should be fixed now, rather than the error case. |
This makes gridPathCells slightly more robust by attempting interpolation from both sides. Previously failing test is now fixed so I added another example pair.