Database Reference
In-Depth Information
You can also build any visualization you want by writing a little bit of custom
code; this may sound like a lot of work but is simpler than it sounds. In
addition, writing your own graphing code lets you customize your graphs
to show exactly what you want to show, not what some tool decides to
show you. Chapter 8, “Putting It Together,” provided sample code to build
dashboards for a data collection app, using Google's charting API, Gviz
time series graphs.
is a general-purpose, open source JavaScript library for turning data into
HTML DOM elements. It can build some impressive graphs and animations,
and there are a number of extensions available for building different types
of charts. That said, it is also more involved and can be a lot of work to set
up even fairly simple graphs.
Summary
This chapter walked you through a number of third-party tools built on top
of BigQuery. Some of these tools, such as Tableau and BIME, enable you
to visualize your data; others, such as the Simba BigQuery ODBC driver,
enable you to integrate BigQuery into your existing software with little or
no changes necessary. The section on Encrypted BigQuery showed how you
can encrypt your data client side, such that Google's servers never see the
unencrypted data.
In addition, this chapter described two different mechanisms for scientific
computing using BigQuery: bigrquery, which lets you run BigQuery queries
from R, and Python pandas, which lets you do the same from Python's
scientific computing environment. Both of these systems are open source, so
if you want to see how they work or reuse any of the components, you can do
so fairly easily.
Along the way, you graphed the relationship between the lengths of
Shakespeare's plays and when he wrote them, built a machine-learning
model to classify Shakespeare plays by genre, and clustered those plays into
histories and other plays.
Finally, you saw only a sampling of the currently available third-party tools
for BigQuery; the SAS connector, visualization options such as QlikView,
