Adding a legend to Iris example.

Hi I was looking at the two implementations of the Iris graph example:

The one that uses bokeh.plotting: http://bokeh.pydata.org/en/latest/docs/gallery/iris.html

and the one that uses bokeh.charts: http://bokeh.pydata.org/en/latest/docs/gallery/iris_scatter_chart.html

Now, I was trying to add a legend to the first graph and I just wanted to ask here whether I am doing it correctly. Here is the code I used:

from bokeh.plotting import figure, show, output_notebook, output_file
from bokeh.sampledata.iris import flowers

colormap = {‘setosa’: ‘red’, ‘versicolor’: ‘green’, ‘virginica’: ‘blue’}
flowers[‘colors’] = [colormap for x in flowers[‘species’]]

p = figure(title = “Iris Morphology”)
p.xaxis.axis_label = ‘Petal Length’
p.yaxis.axis_label = ‘Petal Width’

p.circle(flowers[“petal_length”][flowers[‘colors’]==‘red’], flowers[“petal_width”][flowers[‘colors’]==‘red’],

     color=flowers['colors'][flowers['colors']=='red'], fill_alpha=0.2, size=flowers['sepal_width']*4,legend='Setosa')

p.circle(flowers[“petal_length”][flowers[‘colors’]==‘green’], flowers[“petal_width”][flowers[‘colors’]==‘green’],

     color=flowers['colors'][flowers['colors']=='green'], fill_alpha=0.2, size=flowers['sepal_width']*4,legend='Versicolor')

p.circle(flowers[“petal_length”][flowers[‘colors’]==‘blue’], flowers[“petal_width”][flowers[‘colors’]==‘blue’],

     color=flowers['colors'][flowers['colors']=='blue'], fill_alpha=0.2, size=flowers['sepal_width']*4,legend='Virginica')

output_notebook()
show§

``

That will add one item in the legend for each of the three types of flowers. So, I just wanted to ask whether this is the correct way to add multi-item legends using the plotting interface, or am I missing some a more bokehish method there?

Hello Adi,

First of all I think is great that take current examples and try different things to gain understanding instead of simply copy/paste. Having said that, I believe what you did is the normal way to add legend to a glyph, but we can wait for an expert from the group to confirm this.

I would suggest however using a loop instead of manually setting each group, as other data sets with hundreds of classes would bit quite difficult to process. If you are familiar with pandas you can do something like:

colormap = {‘setosa’: ‘red’, ‘versicolor’: ‘green’, ‘virginica’: ‘blue’}

flowers[‘colors’] = [colormap for x in flowers[‘species’]]

flowers = flowers.set_index(“species”)

p = figure(title = “Iris Morphology”)

p.xaxis.axis_label = ‘Petal Length’

p.yaxis.axis_label = ‘Petal Width’

for variety in flowers.index.unique():

p.circle(flowers.loc[variety,‘petal_length’], flowers.loc[variety,“petal_width”],

color=flowers.loc[variety,“colors”], fill_alpha=0.2, size=10, legend=variety)

show§

``

Best!