.sample incorrect key_dimension values

[holoviz]

.sample incorrect key_dimension values

[philippjfr]

Good catch will have a look.

[philippjfr]

As far as I can tell this was fixed by acd6e45. Try pulling the latest version and let me know if you can still reproduce this bug.

[philippjfr]

That was only concerning the sampling issue. The issue with the axes we addressed previously in #42, but I haven't yet acted on it. Will hopefully get around to that today.

Edit: The easiest workaround for now is to enable the situate_axes plot option on the Image.

[basnijholt]

Yes, that fixed it. I had already installed the latest master but forgot to reload...

Is it possible to plot the same in the same range as the graph that you're sampling? The axis are still from 0 to 35.

[philippjfr]

As I said the situate_axes plot option on Image should do it. The issue is that our entire normalization system currently doesn't allow you to independently control whether the axes ranges or just the Raster luminance values are shared and since it's often pretty annoying for Images to embed in a much larger space we added the situate_axes plot option to make that behavior optional. After the discussions in #49 and #52, we no longer think that is a good idea but have yet to act on the conclusions we've come up with.

Edit: Just noticed that you want both not to be in the 0-35 range? Can't you just not set the Chemical potential Dimension range or am I misunderstanding something?

[basnijholt]

Yes that is a possible solution, but is there no other way?

My understanding was that in Dimension(nane, range(min, max) I put the outer values of that dimension, or am I mistaken?

[philippjfr]

The Dimension range is basically the space that the values along that Dimension lives in. So yes your understanding is correct and as I keep saying Image currently has special treatment in the way it handles those ranges, ignoring them unless situate_axes is enabled. We will get rid of this special handling but for now your options are either not to supply the Dimension.range (based on your plot above that will result in a 5-9 range along the Chemical potential axis) or to enable situate_axes on Image (resulting in a 0-35 range on both the Image and the Curve plot). The behavior you get with situate_axes will eventually be enabled by default.