Make_subplots extract/highlight one subplot

Dash Python
1

We haved created a set of subplots, e.g.:

rows, columns, and list of names populated appropirately

fig = make_subplots(rows=rows, cols=columns, subplot_titles=names)

and then proceed to populate them by adding traces with appropraite row and col keywords.

We would like to extract one of the subplots to use in a new figure. Construction of the plot takes a bit of time and it should be faster to extract a specific row and column from the figure data. We had hoped that the subplot would be relatively self-contained and easy to extract, but it is not obvious that this is the case. Is there a simple way to retrieve a specific subplot and bind it to a Figure that we are returning?

Thanks - Marie

2

Hi @marie, this sounds similar to this thread:

Let us know, if it helps you.

1 Like
3

Thank you @AIMPED, this was helpful indeed. I ended up using an internal method on the figure, _get_subplot_rows_columns to access the geometry and then select_traces with a specific row and column:

   def plot_subfigure(subplot, fig_subplots):
        """
        Retrieve a specific figure subplot and return it as a new figure
        :param subplot:  Index of subplot 1:N
        :param fig_subplots:  Figure with subplots
        :return: New figure
        """

        # Get the number of rows and columns
        row_geom, col_geom = fig_subplots._get_subplot_rows_columns()
        # Column major order, so plots per column will be our divisor
        plots_per_col = len(list(col_geom))

        # Compute row/col for referenced subplot
        subplot = subplot - 1  # 1:N --> 0:N-1
        r = int(subplot / plots_per_col) + 1
        c = subplot % plots_per_col + 1

        # select traces associated with the subplot and add them to new Fig
        traces = []
        generator = self.current_subplots[0].select_traces(row=r, col=c)
        for t in generator:
            traces.append(t)
        fig = go.Figure(data=traces)

        # Update axes - Retrieve axis information for x and y
        # then copy axis information to the new figure's axes
        for upd, get in zip([fig.update_xaxes, fig.update_yaxes],
                            [fig_subplots.select_xaxes, fig_subplots.select_yaxes]):
            ax = [a for a in get(row=r, col=c)][0]
            upd(ax)   # copy properties
            # See if the subplot had a restricted viewport
            if 'domain' in ax:
                upd({'domain': (0, 1)})  # Expand to fill plot

        return fig

Thanks again - Marie

1 Like