You should probably use the recently added json_item for embedding in this case. This function produces a JSON representation of whatever it is passed, that can be embedded separately using a JavaScript call to Bokeh.embed.embed_items in your template.
Here is a complete example based on Flask but the idea should transfer directly to Django:
from __future__ import print_function
import json
from bokeh.embed import json_item
from bokeh.plotting import figure
from bokeh.resources import CDN
from bokeh.sampledata.iris import flowers
from flask import Flask
from jinja2 import Template
app = Flask(__name__)
page = Template("""
<!DOCTYPE html>
<html lang="en">
<head>
{{ resources }}
</head>
<body>
<div id="myplot"></div>
<div id="myplot2"></div>
<script>
fetch('/plot')
.then(function(response) { return response.json(); })
.then(function(item) { Bokeh.embed.embed_item(item); })
</script>
<script>
fetch('/plot2')
.then(function(response) { return response.json(); })
.then(function(item) { Bokeh.embed.embed_item(item, "myplot2"); })
</script>
</body>
""")
colormap = {'setosa': 'red', 'versicolor': 'green', 'virginica': 'blue'}
colors = [colormap[x] for x in flowers['species']]
def make_plot(x, y):
p = figure(title = "Iris Morphology", sizing_mode="fixed", plot_width=400, plot_height=400)
p.xaxis.axis_label = x
p.yaxis.axis_label = y
p.circle(flowers[x], flowers[y], color=colors, fill_alpha=0.2, size=10)
return p
@app.route('/')
def root():
return page.render(resources=CDN.render())
@app.route('/plot')
def plot():
p = make_plot('petal_width', 'petal_length')
return json.dumps(json_item(p, "myplot"))
@app.route('/plot2')
def plot2():
p = make_plot('sepal_width', 'sepal_length')
return json.dumps(json_item(p))
if __name__ == '__main__':
app.run()