New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Axes3D quiver: variable length arrows #4211
Comments
You can also specify the |
Sure, but in that case, shouldn't the normalisation of the I haven't really use masked numpy arrays, but I think the snippet below from
|
Just to confirm, if I turn off the normalisation, things do seem to work as I'd hoped. I don't know whether there would be undesired knock-on effects from this change though. |
That does seem like what it should be doing (as that is how 2D quiver works). |
Can you put in a PR making that change? |
Sure. I'll make one and reference this issue. I'm working on something else at the moment, but should get round to it in a few hours. |
Thanks! Don't stress too much about getting it done right away the deadline for this to make it into the next version is July (you also, don't procrastinate too long). |
Another small thing, where should the arrows be? At the moment, the arrows seem to point from
My feeling is that option 3 is the most intuitive of the three. However, a quick look at StackOverflow suggests that the 2d quiver does option 2 and consistency is probably what we're going for. |
The documentation states that the xyz coordinates are the default locations On Fri, Mar 13, 2015 at 9:25 AM, Kyle Sutherland-Cash <
|
The |
Yes, they should be consistent. Despite my normal insistence on maintaining the API, I think this is an ok On Fri, Mar 13, 2015 at 1:30 PM Kyle Sutherland-Cash <
|
I agree that the normalization should be removed (or maybe made into a keyword argument defaulting to False?). Also, pivoting from the tail seems like the most sensible to match 2D quiver. It seems like this never got changed... Should I put in a PR for this? |
The sooner, the better, as the 2.0 release is getting put together right On Tue, Nov 10, 2015 at 11:18 AM, Danhickstein notifications@github.com
|
Okay, PR created. Hopefully I did it correctly. I'm not too GitHub savvy. :) |
@WeatherGod: As we've been making default changes, for many we've been adding a method to get the old behavior (usually just through |
It would be easy enough to include an option for the normalization. Though it's hard for me to think of a scenario where someone would desire this behavior. |
The use case is when you are only interested in the direction, not the On Tue, Nov 10, 2015, 13:50 Danhickstein notifications@github.com wrote:
|
Yes, I can see that there are some circumstances where the user would want to normalize their data, but should this be built into the quiver function? It's not in the 2D case... |
@mdboom, +1 on having an option to normalize since the feature already On Tue, Nov 10, 2015 at 2:05 PM, Danhickstein notifications@github.com
|
Sounds good. I updated the PR to include a "normalize" keyword argument that defaults to False and controls the behavior. |
Closed by #5458 |
Would there be interest in allowing the
length
argument toAxes3d.quiver()
optionally be an array instead of just a single float?It's a feature I'd like, but I wasn't sure whether to do it properly and make a PR or if there's a good reason for keeping things as is.
The text was updated successfully, but these errors were encountered: