Plotting multiple dataframes in a single output using Python Bokeh - how to create legends for each plot in a loop?

Question

I am trying to input a single file that contains data from different packet streams (hence, different time values). I created a dataframe for each time and the data points/columns from each dataframe are plotted in single plot. I am not being able to get how to append the items in the legend to reflect the correct col names in the loop. This is what I have so far:

j = 0
tabs_df_list = []
## Creating source and figures for each df
for x in df_list:
    j += 1
    df_col_names = x.columns.to_numpy()
    col_names = df_col_names.tolist()
    source = "source"+str(j)
    source = ColumnDataSource(x)
## Create a figure for each df
    p = "p" + str(j)
    p = figure(title = 'Test Tlm Report',
           match_aspect = False,
           toolbar_location = 'right',
           height=750, width=1000,
           x_axis_label = 'Time [hh:mm]', # needs to correct time series
           y_axis_label = 'Data value')    # need to customize to tlm col

    tabs_temp = []
    items_list = []
    k = 1
    while k < len(col_names):
        r = "r" + str(k)
        r = p.line(x=col_names[0], y=col_names[k], source=source, line_width=2, line_color="lightseagreen", name=col_names[k], legend_label=col_names[k])
        r = p.scatter(x=col_names[0], y=col_names[k], source=source)
        items_list.append(col_names[k])
        items_list.append(r)
        tab = [TabPanel(child=p, title=col_names[0])]
        k += 1
   
        
    p.add_tools(HoverTool(tooltips=tooltips))
    legend = Legend(items=[items_list
                          ],location=(3, -25))
    p.add_layout(legend, 'right')
    
    p.legend.click_policy="mute"    
    
    tabs_df = "tabss" + str(j)
    tabs_df = Tabs(tabs=tab)
    tabs_df_list.append(tabs_df)
grid = gridplot (tabs_df_list, ncols=1)
show (grid)

The items list in the legend is what I do not know how to dynamically add to =/

mosc9575 · Accepted Answer · 2024-07-03 21:28:44Z

In my opinion best practice is to define the Legend right after the figure() call and add each legend entry when creating a renderer using legend_item.

Minimal example:

Let's say this is our data:

import pandas as pd

df = pd.DataFrame({"x": [1,2,3], "a": [1,2,3], "b": [2,2,2], "c": [3,2,1]})

Then we can loop over the column names and add this name to the legend in a call like p.line() with legend_label=column_name.

from bokeh.models import Legend
from bokeh.palettes import d3
from bokeh.plotting import figure, show, output_notebook
output_notebook()

p = figure()
legend = Legend(click_policy='mute')
p.add_layout(legend, 'right')
for i, column_name in enumerate(df.columns):
    if column_name == 'x':
        continue
    p.line(df['x'], df[column], legend_label=column_name, color=d3['Category10'][10][i])
show(p)

Output:

Comment:

If you use the same string for a legend label for multiple renderers, they are linked together in the legend und will be muted with just one click.

Collectives™ on Stack Overflow

Plotting multiple dataframes in a single output using Python Bokeh - how to create legends for each plot in a loop?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related