Skip to main content

Reads and writes data stored in StarTable format; and stores table data inmemory as a Pandas data frame for easy manipulation.

Project description



The pdtable Python package offers interfaces to read, write, and manipulate StarTable data.


The pdtable documentation is available at


Demo: see the pdtable_demo notebook or, if you a Jupyter notebook doesn't do it for you, the notebook's paired script.


pdtable is available from

pip install pdtable

and from conda-forge

conda install pdtable -c conda-forge

Data and metadata: storage and access

Table blocks are stored as TableDataFrame objects, which inherit from pandas.DataFrame but include additional, hidden metadata. This hidden metadata contains all the information from Table blocks that does not fit in a classic Pandas dataframe object: table destinations, column units, table origin, etc.

Data in TableDataFrame objects can be accessed and manipulated using the Pandas API as it the object were a vanilla Pandas dataframe, with all the convenience that this entails.

The StarTable-specific metadata hidden in a TableDataFrame's metadata can in principle be accessed directly; however a much more ergonomic interface is offered via a Table facade object, which is a thin wrapper around TableDataFrame. Table also supports some limited data manipulation, though with the advantage of more easily supporting StarTable-specific metadata; for example, easily specifying column units when adding new columns.


Readers and writers are available for CSV and Excel, both as files and as streams. Parsing is efficient and, by default, lenient, though this is readily customized.

Reading can also be filtered early, such that only certain block types or tables with certain names get fully parsed. This can reduce reading time substantially when reading e.g. only a few tables from an otherwise large file or stream.

Directive blocks are parsed by the readers, and presented to the client code for application-specific interpretation.

Import from and export to JSON is also supported.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pdtable-0.0.20.tar.gz (78.3 kB view hashes)

Uploaded source

Built Distribution

pdtable-0.0.20-py3-none-any.whl (98.8 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page