You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Boost.Histogram uses "rank" to provide the number of axes in a histogram. NumPy uses .ndim to get the number of dimensions of a histogram. Moving between histograms and arrays is a little harder due to the different names (I just got caught telling Nino to use .ndim on a histogram). NumPy used to provide np.rank, but it was deprecated in favor of np.ndim in 1.9 (not 1.10, contrary to the 1.18 release notes claim) and removed in 1.18.
Options:
Support .ndim for normal boost-histogram, and .rank for cpp module. One way to do it, and all that.
Add .ndim and keep .rank, for maximum compatibility with both Boost.Histogram and NumPy.
a. Python only
b. Python + cpp module.
Boost.Histogram uses "rank" to provide the number of axes in a histogram. NumPy uses
.ndim
to get the number of dimensions of a histogram. Moving between histograms and arrays is a little harder due to the different names (I just got caught telling Nino to use.ndim
on a histogram). NumPy used to providenp.rank
, but it was deprecated in favor ofnp.ndim
in 1.9 (not 1.10, contrary to the 1.18 release notes claim) and removed in 1.18.Options:
.ndim
for normal boost-histogram, and.rank
for cpp module. One way to do it, and all that..ndim
and keep.rank
, for maximum compatibility with both Boost.Histogram and NumPy.a. Python only
b. Python + cpp module.
@HDembinski, please advise.
The text was updated successfully, but these errors were encountered: