這很有趣，因爲我可以得到hv.Table工作，但hv.Bars拋出以下異常：你可以得到holoviews渲染器（'散景'）。server_doc爲酒吧工作嗎？ NoneType/ColorConverter錯誤
Traceback (most recent call last): File "/app/.heroku/python/lib/python3.6/site-packages/holoviews/plotting/bokeh/util.py", line 120, in mpl_to_bokeh v = colors.ColorConverter.colors.get(v, v) AttributeError: 'NoneType' object has no attribute 'ColorConverter' and File "/app/.heroku/python/lib/python3.6/site-packages/holoviews/core/options.py", line 165, in exit raise AbbreviatedException(etype, value, traceback) holoviews.core.options.AbbreviatedException: AbbreviatedException: AttributeError: 'NoneType' object has no attribute 'ColorConverter'
例外追溯到我在下面的一行代碼： graph = hv.renderer('bokeh').server_doc(df3)
詳情：
從requirements.txt：
bokeh==0.12.6 holoviews==1.8.2
從app.py：
import holoviews as hv
hv.extension('bokeh')
和
def graph_alpha(dataframe):
# function to create a graph for the alpha cash flow dataframe
df1=dataframe
df1['date'] = df1.date.astype('datetime64[ns]')
k_dims = [('date','Date')]
val_dims = [('netincome', 'EBITDA')]
plot_opts = dict(fill_color='#00AA00')
df2 = hv.Table(df1,kdims=k_dims, vdims=val_dims)
df3 = hv.Bars(df1,kdims=k_dims, vdims=val_dims)
#df3 = hv.Bars(df1,kdims=k_dims, vdims=val_dims).opts(plot=plot_opts)
#df3 = hv.Bars(df2).opts(plot=plot_opts)
#df3 = hv.Bars(df2).opts(style=plot_opts)
graph = hv.renderer('bokeh').server_doc(df3)
return graph
對這個有什麼想法？
我翻遍了所有開放和封閉的問題，並沒有看到這一個。 希望我們能弄明白。
乍得