Plotting with bokeh on the Raspberry Pi

05 Mar 2015

Goal is to get plotting with bokeh on the Raspberry Pi.

Install continuum anaconda python distribution
Install bokeh plotting and get a plot server running
be able to run bokeh examples

[UPDATE August 16, 2015] [THIS IS NOT WORKING … REWRITING]

don’t install miniconda, just use stock raspian pytohn

# this takes > 1 hour on a Raspberry Pi Model B !!!
# most of the time is spent in gcc cc1 command
sudo pip install pandas --upgrade 
# after > 1 hour I get an error
# Successfully installed pandas python-dateutil pytz
# OSError: [Errno 39] Directory not empty: '/home/pi/build/pytz'
# I am ignoring this pytz error (something to do with timezone library???)
#
# now i am sidetracked
sudo pip install ipython
# ipython seems ok
#
# this next one may take awhile, started at 1:50pm
sudo pip install bokeh  
# seems ok, can import pandas in python but it gives error
#
#    /usr/local/lib/python2.7/dist-packages/pandas/computation/expressions.py:21: UserWarning: The installed version of numexpr 2.0.1 is not supported in pandas and will be not be used
#    The minimum supported version is 2.1
#
# fixed error with (hopefully does not cause other problems)
sudo pip install numexpr --upgrade

[ORIGINAL POST STARTS HERE]

Install miniconda

pi@pi40 ~ $ wget http://repo.continuum.io/miniconda/Miniconda-3.5.5-Linux-armv6l.sh
pi@pi40 ~ $ bash Miniconda-3.5.5-Linux-armv6l.sh 

Miniconda will now be installed into this location:
/home/pi/miniconda

  - Press ENTER to confirm the location
  - Press CTRL-C to abort the installation
  - Or specify an different location below

[/home/pi/miniconda] >>> 
PREFIX=/home/pi/miniconda
installing: python-2.7.7-0 ...
installing: openssl-1.0.1h-0 ...
installing: pycosat-0.6.1-py27_0 ...
installing: pyyaml-3.11-py27_0 ...
installing: requests-2.3.0-py27_0 ...
installing: yaml-0.1.4-0 ...
installing: conda-3.5.5-py27_0 ...
Python 2.7.7 :: Continuum Analytics, Inc.
creating default environment...
installation finished.
Do you wish the installer to prepend the Miniconda install location
to PATH in your /home/pi/.bashrc ? [yes|no]
[no] >>> yes

Prepending PATH=/home/pi/miniconda/bin to PATH in /home/pi/.bashrc
A backup will be made to: /home/pi/.bashrc-miniconda.bak


For this change to become active, you have to open a new terminal.

Thank you for installing Miniconda!

Make sure numpy and pandas is correct

pip install numpy –upgrade
pip install pandas –upgrade

Install bokeh

In general ‘conda install bokeh’ does not work on Pi, bokeh is not in repo?

pip install bokeh

For a list of required packages, see http://bokeh.pydata.org/en/latest/tutorial/quick_install.html

from: http://bokeh.pydata.org/en/latest/tutorial/quick_install.html
Ideally, you should have the following libraries installed:

NumPy
Flask
Redis
Requests
gevent
gevent-websocket
Pandas

Currently installed with pip

pi@pi40 ~ $ pip freeze
Flask==0.10.1
Flask-Markdown==0.3
Flask-Misaka==0.3.0
Flask-SocketIO==0.5.0
Jinja2==2.7.3
Markdown==2.6
MarkupSafe==0.23
PyYAML==3.11
Pygments==2.0.2
RPi.GPIO==0.5.11
Werkzeug==0.10.1
backports.ssl-match-hostname==3.4.0.2
bokeh==0.8.1
colorama==0.3.3
conda==3.5.5
gevent==1.0.1
gevent-socketio==0.3.6
gevent-websocket==0.9.3
greenlet==0.4.5
ipython==2.1.0
itsdangerous==0.24
misaka==1.0.2
nose==1.3.0
numpy==1.9.2
pandas==0.15.2
psutil==2.2.1
pycosat==0.6.1
pystache==0.5.4
python-dateutil==2.4.1
pytz==2014.10
pyzmq==14.5.0
requests==2.3.0
six==1.9.0
tornado==4.1
websocket==0.2.1
wsgiref==0.1.2

Grab some example plots from bokeh github

The bokeh example plots are at:

https://github.com/bokeh/bokeh/tree/master/examples/plotting/server

Run the bokeh-server

Remember, this make temporary files in directory you run it in.

–ip 0.0.0.0 will redirect the server to your external IP, in my case http://192.168.1.40:5006

bokeh-server –backend=memory –ip 0.0.0.0

pi@pi40 ~/bokeh_examples $ bokeh-server --backend=memory --ip 0.0.0.0
No module named scipy

    Bokeh Server Configuration
    ==========================
    python version : 2.7.8
    bokeh version  : 0.8.1
    listening      : 0.0.0.0:5006
    backend        : memory
    python options : debug:OFF, verbose:OFF, filter-logs:OFF, multi-user:OFF
    js options     : splitjs:OFF, debugjs:OFF
    
/home/pi/miniconda/lib/python2.7/site-packages/bokeh/server/blaze/__init__.py:19: UserWarning: could not import multiuser blaze server No module named blaze.  This is fine if you do not intend to use blaze capabilities in the bokeh server
  warnings.warn(msg)

Example plots I have working

line_animate.py

# The plot server must be running
# Go to http://localhost:5006/bokeh to view this plot

import time

import numpy as np

from bokeh.plotting import *

N = 80

x = np.linspace(0, 4*np.pi, N)
y = np.sin(x)

output_server("line_animate")

p = figure()

p.line(x, y, color="# 3333ee", name="sin")
p.line([0,4*np.pi], [-1, 1], color="# ee3333")

show(p)

renderer = p.select(dict(name="sin"))
ds = renderer[0].data_source

while True:
    for i in np.hstack((np.linspace(1, -1, 100), np.linspace(-1, 1, 100))):
        ds.data["y"] = y * i
        cursession().store_objects(ds)
        time.sleep(0.05)

Now use Flask + bokeh to generate a single html page (no bokeh-server)

eventual goal here is to have Flask server inject new data (via socketio) into html page with bokeh plot by modifying x/y data with javascript (on the client)
simple example that works: https://github.com/bokeh/bokeh/tree/master/examples/embed/simple
more complex: https://github.com/bokeh/bokeh/tree/master/examples/embed/spectrogram

Robert Cudmore archive tags search software