Improve the Readme #87
Conversation
| Return the normalized `Vector2` for the given `Vector2`. | ||
| ### normalize() | ||
|
|
||
| Return a vector with the same direction, and unit length. |
|
@pathunstrom The CI is going to fail because there's one Ideally, I would like |
|
We could also special case multiples of 90deg? I suspect that'll come up with more problems |
|
@astronouth7303 Do you mean that special-casing multiples of 90 degrees will come with more problems, or that solving the general problem will? |
This solves an issue exposed in ppb#87, where imprecisions results in non-null components where the user would expect 0, for instance: >>> Vector2(1, 0).rotate(90) Vector2(6.123233995736766e-17, 1.0)
|
@astronouth7303 Fixed in #89 |
This solves an issue exposed in ppb#87, where imprecisions results in non-null components where the user would expect 0, for instance: >>> Vector2(1, 0).rotate(90) Vector2(6.123233995736766e-17, 1.0)
This solves an issue exposed in ppb#87, where imprecisions results in non-null components where the user would expect 0, for instance: >>> Vector2(1, 0).rotate(90) Vector2(6.123233995736766e-17, 1.0)
|
Could someone rerun the build? |
- scale(1) is equivalent to normalize(), not truncate(1); - truncate is better explained in terms of scale, than the reverse.
- Systematically use Vector2, not Vector. - Use the right floats in return values, not integers or incorrect values.
Also use tuples for the second operands everywhere, as it reduces line-noise.
|
Nevermind, I had to rebase anyway now that #79 was merged, so I could add documentation for vector-scalar division. |
|
How likely is the doctest stuff to break? I'm nervous that Floating Point will happen. |
|
@astronouth7303 Pretty unlikely? Floating point arithmetic is deterministic, so no random failures. A code change can still change the precision at which some values are computed, but the examples use exact values, except for |
scale_to may now overflow with vector lengths larger than 1e75, which is a non-issue.
|
Mmmh, can't brain enough r.n. to find some exact values for |
Vector2.angle.scaleandtruncate.