Awesome

chicago-crimes

Exploring Chicago crimes dataset with DuckDB, Malloy Data, and soon new Panel/PyScript data and dashboard tools ...

Data Source

Data from: https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-Present/ijzp-q8t2

Note: Chicago crimes data is too large for a github repository. You can download it from the the link above.

Raw Data Views

In VSCode

Raw dataset view in VSCode with Tabular Data Viewer and Rainbow CSV:

In DBeaver

Crimes CSV data imported in DBeaver for comparison:

In Tad Viewer

Crimes CSV data imported into Tad Viewer:

With Polars

Quick Chicago crimes CSV data scan and Arrests query with Polars in one cell code block :

With Polars Parquet

Loading Chicago crimes .parquet data file with polars.read_parquet():

With PyArrow

Loading Chicago crimes raw CSV data with PyArrow CSV:

With PyArrow Feather and Parquet

Writing and reading Chicago crimes PyArrow Table data in Feather and Parquet data file formats:

With DuckDB

Loading Chicago crimes CSV data into a blank in-memory DuckDB instance:

With DuckDB Parquet

Loading Chicago crimes .parquet data via DuckDB read_parquet():

With DuckDB SQLMagic

Loading Chicago crimes CSV data into a blank in-memory DuckDB with ipython-sql SQLMagic in VSCode Jupyter Notebook:

With Malloy Data

Loading Chicago crimes 2022 parquet data with Malloy Data tools via DuckDB parquet data table source, with queries, data schema, Malloy queries outline, data preview, and query results displayed in VSCode Malloy extension query editor and views:

With Pandas

Loading Chicago crimes CSV data with Pandas:

In R Studio

Loading Chicago crimes CSV data with DBI R library and DuckDB R API in R Studio:

With Julia REPL

Reading Chicago crimes CSV data with DuckDB Julia Package in VSCode Julia lang, and running it in Julia REPL:

With Julia CSVFiles and DataFrame

Loading Chicago crimes CSV data via Julia CSVFiles into native Julia DataFrames:

In Emacs

Reading Chicago crimes CSV data with SBCL + cl-duckdb in Emacs + SLY:

Visualizations

Collection of Jupyter notebooks and data apps visualizing Chicago crimes data from above.

With Matplotlib

Visualizing Chicago crimes data loaded with Pandas using Matplotlib:

With Altair Charts

2001-2022 Chicago crimes data loaded from a parquet file and summarized with Pandas and Altair charts:

With PyScript

2022 Chicago crimes data loaded from a CSV file with data summary Altair charts in a browser, using Pyodide runtime and Pandas:

https://randomfractals.github.io/chicago-crimes/apps/pyscript/

With Malloy Charts

Displaying Chicago crimes 2022 parquet data with Malloy Charts using Malloy Import with table source, measures, and data queries defined in Malloy Data section above:

With Malloy Composer

View and query 2022 Chicago crime reports data loaded from parquet file with Malloy Composer app in your browser:

https://randomfractals.github.io/chicago-crimes/apps/malloy-composer/

With Malloy Fiddle

View 2022 Chicago crime reports data, schema, Malloy model, and queries with Malloy Fiddle app in your browser:

https://randomfractals.github.io/chicago-crimes/apps/malloy-fiddle

With DuckDB Sql Tools

Loading and querying 7,687,725 Chicago crime reports recorded from 2001 through the end of November 2022 from a large 1.68 GB CSV data file with new VSCode DuckDB Sql Tools extension:

Exporting in-memory DuckDB instance with DuckDB Sql Tools:

Exporting DuckDB instance in Parquet data format and importing it into new test DuckDB memory instance:

Prior Works

Links to our prior works on Chicago Crimes EDA circa 2017/2018:

🔗 Chicago Crimes EDA Summary on Linkedin

📚 Chicago Crimes EDA 2017 Jupyter Notebooks

📚 Observable JS Chicago Crimes Notebook Collection 2018

📚 Chicago Homocides Observable JS Notebook Collection 2018

🐦 Chicago Crimes EDA, Data Preview and Tabular Data Viewer tweets