Fix roundoff issue in Tiles

[dnesbitt61]

Issue

#fixes #4584
Using float in the Tiles Row and Col methods can lead to an incorrect tile id being computed (along with incorrect base lat, lng). Using double as the argument to these methods fixes the issue.
For discussion - will this change cause any compatibility issues?

Tasklist

• Add tests
• Add #fixes with the issue number that this PR addresses
• Update the docs with any new request parameters or changes to behavior described
• Update the changelog
• If you made changes to the lua files, update the taginfo too.

Requirements / Relations

Link any requirements here. Other pull requests this PR is based on?

[gknisely]

@dnesbitt61 need to update the changelog

[kevinkreiser]

im trying to reason about your comment on data compatibilty. old code would expect that a node would land in a different tile than new code. which means old data built with old code would have certain nodes land in the wrong tiles. this would be a problem if we used this method in a scenario where:

1. we had old code
2. we had new data
3. loki used it to find the right tile to look for candidate edges
4. but the n ew data wou ld have the edges in a different tile

the good news is this doesnt really matter because loki never stops looking in one place it looks at closest first and next closes next and so on. the only risk is that in this scenario it finds something that isnt the actual closest thing, which does suck but its not catastrophic. to me i think we can ship this without worrying about compatibility. the data wont crash anything it will at worst give sub par results for these specific edge cases

[dnesbitt61]

how can codecov fail! I added a unit test. Oh well

[nilsnolde]

🎲 🎲

[nilsnolde]

good catch again thanks!