sc.pl.paga_path: TypeError: float() argument must be a string or a number, not 'csr_matrix'

I was plotting the paga path, I got the error of TypeError: float() argument must be a string or a number, not 'csr_matrix'. I guess this might be related with the sparse format of adata.raw.X, because my codes works if I deleted adata.raw. What would be the solution? Thank you

_, axs = pl.subplots(ncols=3, figsize=(6, 2.5), gridspec_kw={'wspace': 0.05, 'left': 0.12})
pl.subplots_adjust(left=0.05, right=0.98, top=0.82, bottom=0.2)
for ipath, (descr, path) in enumerate(paths):
    _, data = sc.pl.paga_path(
        adata, path, gene_names,
        show_node_names=False,
        ax=axs[ipath],
        ytick_fontsize=8,
        left_margin=0.15,
        n_avg=50,
        annotations=['distance'],
        show_yticks=True if ipath==0 else False,
        show_colorbar=False,
        color_map='Greys',
        groups_key='clusters',
        color_maps_annotations={'distance': 'viridis'},
        title='{} path'.format(descr),
        return_data=True,
        show=False)
    data.to_csv('./write/paga_path_{}.csv'.format(descr))
pl.savefig('./figures/paga_path_panglao.pdf')
pl.show()

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
TypeError: float() argument must be a string or a number, not 'csr_matrix'

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
<ipython-input-8-86ecf06e6589> in <module>()
     18         title='{} path'.format(descr),
     19         return_data=True,
---> 20         show=False)
     21     data.to_csv('./write/paga_path_{}.csv'.format(descr))
     22 pl.savefig('./figures/paga_path_panglao.pdf')

5 frames
<__array_function__ internals> in cumsum(*args, **kwargs)

/usr/local/lib/python3.6/dist-packages/numpy/core/fromnumeric.py in _wrapit(obj, method, *args, **kwds)
     45     except AttributeError:
     46         wrap = None
---> 47     result = getattr(asarray(obj), method)(*args, **kwds)
     48     if wrap:
     49         if not isinstance(result, mu.ndarray):

ValueError: setting an array element with a sequence.

Versions:

scanpy==1.5.1 anndata==0.7.3 umap==0.4.4 numpy==1.18.5 scipy==1.4.1 pandas==1.0.5 scikit-learn==0.22.2.post1 statsmodels==0.10.2 python-igraph==0.8.2 leidenalg==0.8.1

Start with the sc.pl.paga_path entry point and reproduce the minimal example using an AnnData object whose adata.raw.X is a scipy csr_matrix. Trace the failing cumsum and array conversion shown in the traceback; done means paga_path completes with sparse adata.raw.X and still returns the requested path data.