same notebook gives different stream plots

[crazyhottommy]

Hi,
I installed STREAM1.0 on a different computer via conda. I ran the same notebook but the stream plots are different. I am wondering what could be the reason?

Thanks!

[crazyhottommy]

also, I am getting this error with the same notebook on my other computer:

st.plot_stream(adata,root='S2',color=['FCGR3A','PRF1'], factor_num_win=5)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/var/folders/2p/3c16s7dd5y9cmg7k8gkb9x800000gp/T/ipykernel_14386/2839065876.py in <module>
----> 1 st.plot_stream(adata,root='S2',color=['FCGR3A','PRF1'], factor_num_win=5)

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/stream/core.py in plot_stream(adata, root, color, preference, dist_scale, factor_num_win, factor_min_win, factor_width, factor_nrow, factor_ncol, log_scale, factor_zoomin, fig_size, fig_legend_order, fig_legend_ncol, vmin, vmax, pad, w_pad, h_pad, save_fig, fig_path, fig_format)
   3138                                    factor_num_win=factor_num_win,factor_min_win=factor_min_win,factor_width=factor_width,
   3139                                    factor_nrow=factor_nrow,factor_ncol=factor_ncol,
-> 3140                                    log_scale=log_scale,factor_zoomin=factor_zoomin)     
   3141         dict_plot['numeric'] = [verts,extent,ann_order,dict_ann_df,dict_im_array]
   3142 

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/stream/extra.py in cal_stream_polygon_numeric(adata, dict_ann, root, preference, dist_scale, factor_num_win, factor_min_win, factor_width, factor_nrow, factor_ncol, log_scale, factor_zoomin)
   1514         if x in nx.get_edge_attributes(flat_tree,'id').values():
   1515             id_cells = np.where(df_stream['branch_id']==x)[0]
-> 1516             df_stream.loc[df_stream.index[id_cells],'edge'] = [x]
   1517             df_stream.loc[df_stream.index[id_cells],'lam_ordered'] = df_stream.loc[df_stream.index[id_cells],'branch_lam']
   1518         else:

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/indexing.py in __setitem__(self, key, value)
    721 
    722         iloc = self if self.name == "iloc" else self.obj.iloc
--> 723         iloc._setitem_with_indexer(indexer, value, self.name)
    724 
    725     def _validate_key(self, key, axis: int):

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/indexing.py in _setitem_with_indexer(self, indexer, value, name)
   1728         if take_split_path:
   1729             # We have to operate column-wise
-> 1730             self._setitem_with_indexer_split_path(indexer, value, name)
   1731         else:
   1732             self._setitem_single_block(indexer, value, name)

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/indexing.py in _setitem_with_indexer_split_path(self, indexer, value, name)
   1767 
   1768             elif np.ndim(value) == 2:
-> 1769                 self._setitem_with_indexer_2d_value(indexer, value)
   1770 
   1771             elif len(ilocs) == 1 and lplane_indexer == len(value) and not is_scalar(pi):

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/indexing.py in _setitem_with_indexer_2d_value(self, indexer, value)
   1828         if len(ilocs) != value.shape[1]:
   1829             raise ValueError(
-> 1830                 "Must have equal len keys and value when setting with an ndarray"
   1831             )
   1832 

ValueError: Must have equal len keys and value when setting with an ndarray

[crazyhottommy]

any help on this?

[huidongchen]

Hi Ming, sorry about the delay. It has been quite hectic lately.

This seems an issue related to the version of pandas #115 . Can you try to downgrade pandas to v1.0 and see if it works for you? (I noticed this did not solve a similar issue mentioned by another user but it's still worth a try for now)

[crazyhottommy]

Yeah. the pandas version is tricky.
I downgraded it to v1.0, now a different error

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
/var/folders/2p/3c16s7dd5y9cmg7k8gkb9x800000gp/T/ipykernel_59461/2817862961.py in <module>
      1 st.plot_stream_sc(adata,root='S2',color=['label','FCGR3A', 'PRF1'],
----> 2                   dist_scale=0.3,show_graph=True,show_text=True)

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/stream/core.py in plot_stream_sc(adata, root, color, dist_scale, dist_pctl, preference, fig_size, fig_legend_ncol, fig_legend_order, vmin, vmax, alpha, pad, w_pad, h_pad, show_text, show_graph, save_fig, fig_path, fig_format, plotly)
   2894         raise ValueError("There is no root '%s'" % root)
   2895 
-> 2896     add_stream_sc_pos(adata,root=root,dist_scale=dist_scale,dist_pctl=dist_pctl,preference=preference)
   2897     stream_nodes = adata.uns['stream_'+root]['nodes']
   2898     stream_edges = adata.uns['stream_'+root]['edges']

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/stream/extra.py in add_stream_sc_pos(adata, root, dist_scale, dist_pctl, preference)
    398     dict_edge_shift_dist = calculate_shift_distance(adata,root=root,dist_pctl=dist_pctl,preference=preference)
    399     dict_path_len = nx.shortest_path_length(flat_tree,source=root_node,weight='len')
--> 400     df_cells_pos = pd.DataFrame(index=adata.obs.index,columns=['cells_pos'])
    401     dict_edge_pos = {}
    402     dict_node_pos = {}

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/frame.py in __init__(self, data, index, columns, dtype, copy)
    433             )
    434         elif isinstance(data, dict):
--> 435             mgr = init_dict(data, index, columns, dtype=dtype)
    436         elif isinstance(data, ma.MaskedArray):
    437             import numpy.ma.mrecords as mrecords

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/internals/construction.py in init_dict(data, index, columns, dtype)
    237             else:
    238                 nan_dtype = dtype
--> 239             val = construct_1d_arraylike_from_scalar(np.nan, len(index), nan_dtype)
    240             arrays.loc[missing] = [val] * missing.sum()
    241 

/opt/anaconda3/envs/stream1.0/lib/python3.7/site-packages/pandas/core/dtypes/cast.py in construct_1d_arraylike_from_scalar(value, length, dtype)
   1447     else:
   1448         if not isinstance(dtype, (np.dtype, type(np.dtype))):
-> 1449             dtype = dtype.dtype
   1450 
   1451         if length and is_integer_dtype(dtype) and isna(value):

AttributeError: type object 'object' has no attribute 'dtype'

[crazyhottommy]

downgrade to v1.0.5 solved the problem.

[huidongchen]

Thanks Ming for the feedback. I'm glad you found a solution.

I will keep the issue open for now. I'm hoping to release a quick fix in the next few days.

[crazyhottommy]

is stream 1.1 on conda?

[huidongchen]

Hi Ming, yes, it will be available on bioconda soon! bioconda/bioconda-recipes#32071

[huidongchen]

stream v1.1 is available now. The issues caused by the latest version of pandas have been fixed in the new version.