hshhrr / plotly-upset Goto Github PK

Hey, it seems that when sorting along the y-axis is on, the order of the black dots and edges is not correct anymore. It seems that reordering of the dots and edges is not implemented for sorted_y (see difference upper and lower). Could you perhaps take a look at it? Otherwise the package is absolutely awesome

 if sorted_x is not None:
            a, b, c = int_ss, string_repr, np.arange(0, 2 ** len(sets))
            order = False if sorted_x.lower() == "a" or sorted_x.lower() == "ascending" else True
            sorted_list = sorted(zip(a, b, c), reverse=order)
            transposed = np.array(sorted_list).T

            a, b, c = transposed

            int_ss, string_repr, _sorted_sequence = a.astype(int), b, c.astype(int)

            t, f, edges = get_sorted_nodes_and_edges(
                t=t, f=f, edges=edges,
                sorted_sequence=_sorted_sequence
            )

        if sorted_y is not None:
            a, b = ind_ss, sets
            order = False if sorted_y.lower() == "a" or sorted_x.lower() == "ascending" else True
            sorted_list = sorted(zip(a, b), reverse=order)
            transposed = np.array(sorted_list).T
            a, b = transposed
            ind_ss, sets = a.astype(int), b

Hi:

Thank you for such a great library. This is more a question rather than an issue: I have a dataset with many groups (more than 5), meaning that the combination, or intersections, is quite large, and hence I was wondering if it is possible to generate the bar plot with horizontal bars and also with rotation of the connected dots to grow the list of combinations vertically and not horizontally.

Again, thank you for your package.

I really like to use plotly and am currently experimenting more complex visualisations. Sofar, I really like plotly_upset. As I really prefer to use polars over pandas, I would like to add support for polars.DataFrame to this project.