Rebinning functionality using a new axis

[MoAly98]

Hi experts,

I was wondering if it is possible to implement a rebinning algorithm that uses a new bh.axis or list of new edges object to complement the rebin-by-factor method that exists? Maybe one can start by an implementation that asserts the new edges lie on the boundaries of the old ones and then work on adding some interpolation algorithms to handle other cases. I have had a simple implementation of the re-binning which looks like :

def rebin(hist, new_axis, axis = 0):
    # Need to match storage of incoming histogram
    hnew = bh.Histogram(new_axis, storage=bh.storage.Weight())
    new_edges = new_axis.edges
    sw =  np.array(hist.values())
    sw2 =  np.array(hist.variances())
    assert all(e in hist.axes[axis].edges for e in new_edges)
    binmap = np.digitize(hist.axes[axis].centers, new_edges)
    new_sw, new_sw2 = [], []

    for bin_num in sorted(set(binmap)):
        old_bins_in_this_bin = np.where(binmap == bin_num)[0]
        w_slice = sw[old_bins_in_this_bin].sum()
        sw2_slice = sw2[old_bins_in_this_bin].sum()
        new_sw.append(w_slice)
        new_sw2.append(sw2_slice)
    hnew[...] = np.stack([new_sw, new_sw2], axis=-1)

    return hnew

Or, it would be simpler if one can slice with an array of bh.loc(new_edges) for example.

Has there been any discussion/issues created on this?