MyPythonDB

Python is at the heart of most data analysis but to share analysis's most users have been uploading data to SQL databases, until now. MyPythonDB turns your python instance into an SQL database. Existing tools can query data directly at faster than database speeds.

• PythonDB provides a MySQL interface that allows every BI tool to just work.
• With DuckDB and Polars-SQL you can directly query in-memory dataframes, S3 or on-disk parquet files using SQL.
• Via the simple SQL editor web interface, you can share links to results with colleagues.

Modern tools and libraries are arriving and providing capabilities and performance only previously available accessible by paying for huge enterprise licenses $$$. Given these changes, We believe data analysts will increasingly be performed using these tools in future.

Uses

1. Import package to expose your python instance as a MySQL Database.
2. pythondb.exe mydb.duckdb - Load a duckdb database and make it remotely accessible.
3. pythondb --language duckdb - A new in-memory duckdb instance
4. pythondb code.py --language polars - A polars instance

Python as MySQL

import mypythondb

mypythondb.start(port=3145, webport=9090, language='POLARS')

Command Line Options

Options:
  -l, --language LANG    Language to interpret code as.  [default: PYTHON]
  -c, --command COMMAND  Run COMMAND
  -P, --port SQLPORT     Port for MySQL compatible server to listen on
  -w, --webport WEBPORT  Port for webserver to listen on
  -q, --quiet            Quiet, don't show banner
  -v, --verbose          Display debugging information
  --version              Show the version and exit.
  --help                 Show this message and exit.

Development Info

Facts

• Python is the language for data analysis.
• Polars / DuckDB provides extremely fast DataFrame operations previously only accessible at great cost.
• Many libraries and databases are converging on common data formats
  • Apache Arrow in memory
  • Apache Parquet on disk

Commands

poetry install --sync
poetry lock
poetry run pythondb
poetry run pytest
poetry run pyinstaller --onefile  launcher.py

Optional TODO

• Ability to set duckdb conn to one of users choosing.

• Ability to set polars context to one of users choosing.

• QWEB

  • autocomplete for duckdb / polars / python?
  • Sharing link to show correct chart type fix.
  • Server Tree Working

• MySQL

  • Test with tableau / dbeaver / ?
  • Server Tree Working
  • Security - Allow setting authentication function

• Tutorials/Website

  • Using QDB to enable DuckDB IPC
  • Using QDB to query python direct from tableau / every popular combo......
  • Using QDB to enable polars SQL IPC

• Deploy

  • Provide mac built package
  • Provide linux built package
  • Get into linux package managers
  • Get into mac package managers

• KDB

  • Subscribe to streaming data
  • KDB Server ability to allow users to send data in from a kdb+ process?
  • Q Language?

• Misc

  • K Language?
  • Other languages?