Extract selected data in Bokeh to text file and other questions

Question

I am currently working on a Bokeh script that will display an earthquake catalogue on a map. So far, I have had little trouble, but I have stumbled upon a couple questions.

1) I came upon this stackoverflow Q&A: Get selected data contained within box select tool in Bokeh, which was very helpful in learning how to use CutsomJS to take the selected data and output it. I have, however encountered a bit of oddness. The script only seems to work properly for me in the Jupyter notebook. The CustomJS portion reads as follows:

source.callback = CustomJS(args=dict(p=p), code="""
        var inds = cb_obj.get('selected')['1d'].indices;
        var d1 = cb_obj.get('data');
        console.log(d1)
        var kernel = IPython.notebook.kernel;
        IPython.notebook.kernel.execute("inds = " + inds);

If I try running it through and html window, the following command has no output:

subset = zip([source.data['xvals'][i] for i in inds],
    [source.data['yvals'][i] for i in inds])
subset_array = np.array(subset)
outfile = 'Scripts/subset.xy'
fo = open(outfile,'w')
for i in range(len(subset_array)):
    print(subset_array[i,0],subset_array[i,1],file=fo)
fo.close()

I could use some advice on how to get the selected data to output to a variable, and ultimately a text file when using html, as opposed to the Jupyter notebook.

2) I have been trying to create a legend for my plot. The earthquakes are indicated by circles (so it is a circle plot), and the circles vary in size based on their magnitude and in colour based on their depth, which I have gotten to work. What I would like to do is show circles of different sizes in the legend, each reflecting different magnitudes (Mag 1, Mag 2, Mag 3, etc.). Here's the command I use for plotting the data:

p.circle("xvals", "yvals",source=source, size="radius", fill_color=transform("depth", mapper),legend = 'Magnitude')

I've tried messing around with the legend commands, but it's all a bit difficult for me to grok, even though I feel like this should be a pretty simple thing to do.

I am utilizing the latest version of Bokeh (0.12.7) and Python 2.7.12. Thanks for any help, and I appreciate the time!

score 1 · Answer 1 · answered Sep 10 '17 at 17:57

On question 1):

If you are creating your html file with Bokeh's output_file function, you cannot access any python code from it, since the resulting file is just html + javascript. The file saving functionality could be embedded in the javascript code (a bit tricky since it seems there is no solution that works in every browser). Another option would be to run a bokeh server and implement some mechanism to push the data to the client.

Below is a rough example where the file saving functionality is included in the html file. This should work on non-IE browsers.

from random import random
from bokeh.plotting import figure, output_file, show, ColumnDataSource
from bokeh.models import CustomJS, Button
from bokeh.layouts import row, column

# Create some random data and put it into a ColumnDataSource
x = [random() for x in range(100)]
y = [random() for y in range(100)]
source_data = ColumnDataSource(data=dict(x=x, y=y))

# Create a button that saves the coordinates of selected data points to a file
savebutton = Button(label="Save", button_type="success")
savebutton.callback = CustomJS(args=dict(source_data=source_data), code="""
        var inds = source_data.selected['1d'].indices;
        var data = source_data.data;
        var out = "";
        for (i = 0; i < inds.length; i++) {
            out += "(" + data['x'][inds[i]] + ", " + data['y'][inds[i]] +") ";
        }
        var file = new Blob([out], {type: 'text/plain'});
        var elem = window.document.createElement('a');
        elem.href = window.URL.createObjectURL(file);
        elem.download = 'selected-data.txt';
        document.body.appendChild(elem);
        elem.click();
        document.body.removeChild(elem);
        """)

# Plot the data and save the html file
p = figure(plot_width=400, plot_height=400, tools="lasso_select, reset")
p.circle(x='x', y='y', source=source_data)
plot = column(p, savebutton)
output_file("test.html")
show(plot)

Extract selected data in Bokeh to text file and other questions

1 Answers1