Computing Conditional Quantiles

Python usage
1

The method computeConditionalQuantile of a distribution seams only to deal with the last component. What is the trick to get the conditional quantile of say X2 of a 3 dim distribution X1,X2,X3 ie. the conditional quantile of the second component, not the last.

2

One possible solution is to reorder the components using getMarginal (here swap x1 and x2):

dist = ot.Normal([10,0], ot.CovarianceMatrix([[1.0, 0.5], [0.5, 1.0]]))
dist.getMarginal([1,0]).computeConditionalQuantile(0.9, [0])
3

Hi!
The ConditionalDistribution in the otbenchmark module can be used to create the conditional distribution. Then the quantile can be computed once the distribution is defined.
Regards,
Michaël

4

careful there is already a ConditionalDistribution in OT, but its a different thing: the parameters are conditioned by another distribution