CustomJSTransform on Annotation

[gmerritt123]

Hi. I’m wondering if there is currently a way to drive the parameters of an annotation via CustomJSTransform.

Something like:

main_src = ColumnDataSource({'x':[1,2,3,4]})
#transform that calcs the average x of src
tr = CustomJSTransform(args=dict(src=src)
                       ,func='''
                       return src.data['x'].reduce((a, b) => a + b, 0) / src.data['x'].length;
                       '''
                       )
sl = Slope(gradient=tr # or transform(field_name='dum',transform=tr) ?? also doesn't work
           ,y_intercept=0, #etc.
           )

Am I missing something obvious? The gradient runs off a singular value (I think that means “non vectorized”), and the transform returns a singular value, but I’m definitely missing something in how to assign the transform to it (or if that is even possible)…

Thanks!

[Bryan]

I don’t think it’s currently possible, AFAICT the gradient property just accepts a number. With the move towards glyph-ification ^[1] this seems like a reasonable thing to support so I’d suggest making a GitHub Issue around supporting transforms for non-vectorized properties. cc @mateusz

---

1. goal: “annotation” is something you do with standard objects, not a different class of objects ↩︎