quick-start

bokeh can provide interactive visualization for modern web browsers. elegant, concise, vesatile graphics, *afford high-performacne interactiivity onver lagerge or streaming datasets.

bokeh provide two interface levels:

1. bokeh.models: low-level interface provide most flexibility
2. bokeh.plotting: higher-level interface centered around composing visual glyphs.

this focus on bokeh.plotting.

installation

# recommend way
# conda install deps, and exmaples in *examples/* subdirectory
conda install bokeh

# using pip, you need install ~numpy~ and so on.
pip install bokeh

get-starting

• [] TODO output bokeh plot to or-babel result

you can download sample-date and example from bokeh

bokeh sampledata

the general steps of bokeh.ploting:

1. prepare data
2. tell bokeh where to generate output(file or notebook)
3. call figure() to create a plot figure
4. add renderers, to add renderers to figure, such as fig.line()
5. call show() or save() to output results

from bokeh.plotting import figure, output_file, show, output_notebook

# prepare some data
x = [1, 2, 3, 4, 5]
y = [6, 7, 2, 4, 5]

# output to static html file
# ??? how to use in web-server like flask.
# output_file("lines.html")
# output_notebook()

# create a new plot with a litter and axis labels
p = figure(title="simple line example", x_axis_label="x", y_axis_label="y")

# add a line renderer with legend and line thickness
p.line(x, y, legend_label="Temp.", line_width=2)

# show the results
show(p)

bokeh work well with jupyter-notebook. jupyter-notebook is the common tool for exploratory data analysis, widely used across the pyData community. you can view online Bokeh NBViewer Gallery, or examples/howto in the Bokeh Repo.

TODO notebook:

1. bokeh interact with jupyter’s dropdown and sliders
2. use Numba to efficiently perform image processing

bokeh notebook cannot display in Github preveiw, cause bokeh using javascript for display which was scrub by Github.

concept

Plot

central concept in Bokeh, containers that hold all the various objects (renderers, guides, data, and tools)

??? the figure() return object

Glyph

basic visual marks that bokeh can display. ??? the line() plot() returns

1. at low-level, there are glyph objects, such as Line on the bokeh.models interface
2. at higher-level, there are glyph methods such as line() provided by bokeh.plotting interface.

Guides and Annotations

tools component that aid presentation or help user make comparisions.

Guide

visual aids help user judge distances, angles etc. such as grid-lines(珊格线) or bands(???布林线), axes(轴线, linear, log, datetime)

annotations

visual aids that label or name parts of the plot. such as titles, legends(图例) etc.

Ranges

the data-space bounds of a plot.

by default of bokeh.plotting interface, it give DataRange1d that try to automatically set the plot bounds to encompass all the available data.

p = figure(x_range=[0, 10], y_range=(10, 20))

Resources

the generate bokeh output, such as html-file or notebook.

bokeh default load bokehjs from cdn.bokeh.org, you can embedeed it static by output_file(mode="inline")

TODO more-exampels

Vectorized colors and sizes

Linked Panning and Brushing

DateTiem axes

TODO Bokeh applications

check-over this run bokeh server

bokeh comes with an optional server-component, the Bokeh Serevr.

it affords many novel and powerful capabilities:

1. UI widgets and plot selections driving computations and plot updates.
2. Intelligent server-side downsampling of large datasets.
3. Streaming data automatically updating plots.
4. Sophisticated glyph re-writing and transformations for “Big Data”.
5. Plot and dashboard publishing for wider audiences.