the interpolation= argument to quantile was renamed to method=, which has additional options.
Users of the modes 'nearest', 'lower', 'higher', or 'midpoint' are encouraged to review the method they used. (Deprecated NumPy 1.22)
Since method= still supports 'nearest' the fix is as simple as replacing interpolation= with method=
Although this only results in a warning, it's a simple fix and would be nice not to have the warning :)
https://github.com/tensorflow/lattice/blob/master/tensorflow_lattice/python/premade_lib.py#L1474
the
interpolation=argument to quantile was renamed tomethod=, which has additional options. Users of the modes 'nearest', 'lower', 'higher', or 'midpoint' are encouraged to review the method they used. (Deprecated NumPy 1.22)Since
method=still supports 'nearest' the fix is as simple as replacinginterpolation=withmethod=Although this only results in a warning, it's a simple fix and would be nice not to have the warning :)