Hover tool question.

Hi. It is not clear why the hover tool does not work in the following example. Could you say, what’s wrong with it?

output_notebook()
TOOLS="pan,wheel_zoom,box_zoom,reset,previewsave,hover"
source = ColumnDataSource(
    data=dict(
        x=np.array(AAPL['date'], 'M64'),
        y=AAPL['adj_close']
    )
)
line(np.array(AAPL['date'], 'M64'), AAPL['adj_close'], color='#A6CEE3', legend='AAPL', source=source,tools=TOOLS)
hover = [t for t in curplot().tools if isinstance(t, HoverTool)][0]
hover.tooltips = OrderedDict([
    ("(x,y)", "($x, $y)")
])
show()

Is it a good place for such questions, or I may ask somewhere else? Thank you.

[chdoig]

Hi! In the documentation you can find the following information on the HoverTool: http://bokeh.pydata.org/docs/user_guide.html#hovertool

Hover tool currently does not work with lines, images, or patch type glyphs.

My suggestion would be to use scatter() + line() (remember to include hold()), since scatter does support hover:

output_notebook()
hold()
TOOLS="pan,wheel_zoom,box_zoom,reset,previewsave,hover"
source = ColumnDataSource(
    data=dict(
        x= np.array(AAPL['date'], 'M64'),
        y=AAPL['adj_close'],
        date = AAPL['date']
    )
)
scatter('x', 'y', color='#A6CEE3', source=source, tools=TOOLS, x_axis_type = "datetime")
line('x', 'y', color='#A6CEE3', legend='AAPL', source=source)
hover = [t for t in curplot().tools if isinstance(t, HoverTool)][0]
hover.tooltips = OrderedDict([
    ("date", "@date"),
    ("price", "$y" )
])
show()

I’ve added some other suggestions:

• If you are passing the source, you don’t need to write again np.array(AAPL['date'], 'M64'), just the name of the column in your ColumnDataSource that you want to use (e.g. 'x').
• If you want to display the date with a nice date format and not the timestamp, you can do date = AAPL[‘date’] and display that in the hover box.
• You can also add a x_axis_type = "datetime".

Hope this helps! Please, let me know if the above code doesn’t work for you.

[hackermd]

Thank you for your feedback! My images are pretty big so that I anticipate problems with the number of points. Since the image data is simply stored as an array I thought it should be straight forward to read out the values from the plot object!? For now I will try to do this on the client side via the generated canvas element.