I'd like to join the others in saying thanks for such a great toolbox!
This is more of a question/ suggestion:
When I run a live update using:
self._live_fplt[0].candlestick_ochl(df_update[['open', 'close', 'high', 'low']], ax=self.ax)
and df_update only contains a subset of indices of the original plot, then PandasDataSource. update first merges both files with
output_df = pd.merge(input_df, df[[]], how='outer', left_index=True, right_index=True)
and then a couple of lines further down, cuts everything away again with
output_df = output_df.loc[input_df.index[start_idx:end_idx], :]
Is this intendend behavior? For me taking all lines of input_df where they exist and filling the rest with df would have been the expected behavior. I could then just "send" the last couple of candles and would get an up to date plot.
I achieved this by using output_df = input_df.combine_first(df) instead of the pd.merge, and by commenting out the cut out data.
If I misunderstood the intention then feel free to just close my comment.
Cheers, H
I'd like to join the others in saying thanks for such a great toolbox!
This is more of a question/ suggestion:
When I run a live update using:
self._live_fplt[0].candlestick_ochl(df_update[['open', 'close', 'high', 'low']], ax=self.ax)
and df_update only contains a subset of indices of the original plot, then PandasDataSource. update first merges both files with
output_df = pd.merge(input_df, df[[]], how='outer', left_index=True, right_index=True)
and then a couple of lines further down, cuts everything away again with
output_df = output_df.loc[input_df.index[start_idx:end_idx], :]
Is this intendend behavior? For me taking all lines of input_df where they exist and filling the rest with df would have been the expected behavior. I could then just "send" the last couple of candles and would get an up to date plot.
I achieved this by using output_df = input_df.combine_first(df) instead of the pd.merge, and by commenting out the cut out data.
If I misunderstood the intention then feel free to just close my comment.
Cheers, H