[Python] How do I make a mark_rule that stops at a previously defined line?

I've plotted a curve which involved some transform calculations, and now I want to place a vertical rule at a specific x location that stops at the height corresponding to that line.



This is what I've managed to do so far, but feels to cumbersome and it does not work entirely either. I'm filtering the value in the x domain that's closest to where I want to place the rule, as follows:

import numpy as np
import altair as alt
import scipy

np.random.seed(1)
_data = pl.DataFrame({"x": scipy.stats.norm().rvs(200)})
_cdf = (
alt.Chart(_data).mark_area(opacity=0.7)
.transform_quantile(quantile="x", as_=["prob", "value"])
.encode(
alt.X("value:Q"),
alt.Y("prob:Q"),
)
)

_rule_x = 0
_rule = _cdf.transform_calculate(
dist=f"abs(datum.value - {_rule_x})"
) .transform_joinaggregate(
min="min(dist)"
).transform_filter(
"datum.dist <= datum.min"
).transform_window(
rank_dist="rank(datum.dist)",
frame=[None, None],
).transform_filter(
"datum.rank_dist <= 5"
).mark_rule().encode(
alt.X("value:Q"),
# alt.XDatum(_rule_x),
alt.Y("prob:Q")
)

_cdf + _rule


The thing is:

1. As you can see in the image above, the rule is not exactly at 0, it's at the value in "x" that's closest to 0.
2. I can force the rule to be at 0 (see the commented line in the code), but then it will not stop at the height of the curve at 0, which is actually a result of interpolation.

I'd like to be able to place the mark_rule exactly wherever I want, and for it to stop at the height defined by the interpolated curve.

Also it feels like I'm doing so many transforms for a relatively simple concept, is there a better way to go about it?

Continue reading...