Different anchor points for measurement data points corresponding to different output variables

[gnodking7]

Hello,

I'm working with a single time-dependent PDE problem with two outputs Q(x,t) and S(x,t) to learn.
I have some empirical/measurement data points for both Q(x,t) and S(x,t) that are to be included in the training process,
but I would like to do this such that I use different spatio-temporal locations (different anchor points) for Q(x,t) and S(x,t).

I have no problem with training using Q(x,t), S(x,t) measurement data points at the same spatio-temporal locations.
In this case, I use X to denote the locations shared by both Q(x,t), S(x,t) measurements, and I use the following codes:

observe_Q = dde.icbc.PointSetBC(X, Qvals, component=1)
observe_S = dde.icbc.PointSetBC(X, Svals, component=0)

data = dde.data.TimePDE(geomtime, pde, [observe_S, observe_Q], num_domain=3540, num_boundary=160, num_initial=320, anchors=X)

But how to do this for two different spatio-temporal locations, say X and Y?
I've tried doing the following, but I get an error that my implementation with anchors=[X, Y] is incorrect.

observe_Q = dde.icbc.PointSetBC(X, Qvals, component=1)
observe_S = dde.icbc.PointSetBC(Y, Svals, component=0)

data = dde.data.TimePDE(geomtime, pde, [observe_S, observe_Q], num_domain=3540, num_boundary=160, num_initial=320, anchors=[Y, X])

[lululxvi]

You don't need anchors for PointSetBC. Everything else looks good.