Bug in `interpolate()` when graph has no `mr` attribute about pygsp HOT 2 OPEN

Thanks for reporting. Can you share the code you wrote that triggered the issue? I'm not familiar with this part of the code, but the mr attribute appears to be created by the pygsp.reduction.graph_multiresolution function, which I would assume you'd call before pygsp.reduction.interpolate.

from pygsp.

Yes, sure. And I'm sorry for the huge delay, I had to step back from this problem and just recently returned to it.

I have a graph signal stored in the 1D array s. I decimated half of its samples and I'm trying to estimate the values in the now "unknown" vertices:

from pygsp.reduction import interpolate
from pygsp import graphs
import numpy as np
import copy

frac_zero = 0.5
s_ = copy.deepcopy(s)
idx_zero = np.random.permutation(len(s))[:int(frac_zero*len(s))]
s_[idx_zero] = 0
bool_is_zero = s_ == 0

# Interpolated signal
s_intep = interpolate(
    G, f_subsampled=s[~bool_is_zero], keep_inds=np.where(~bool_is_zero)[0],
    order=100, reg_eps=0.005)

from pygsp.