JonasHeylen

@JonasHeylen·8 public vizzes

Gist 66b1a9a31dce76682837c3fd4d9a468b

This example demonstrates how to query JSONB data directly in PostgreSQL and flatten it into a relational result set. The code creates a table with a JSONB column, inserts two sample rows, and then uses two different query approaches: the `->>` operator to extract scalar fields as text, and the `jsonb_to_record` function with a column definition list to cast JSON values into typed columns (text and integer). The example highlights both methods for accessing nested JSON data, showing the explicit extraction syntax versus the more structured record-expansion technique, all within a single SQL file.

Feb 9, 2017

Loading thumbnail…

Gist 6e70c835e46d4292e11d165fd0f22c77

This example visualizes the number of business days per month in Belgium for 2017, using a custom holiday calendar to exclude public holidays. It shows how many working days each month contains, revealing monthly variation. The code relies on pandas’ `CustomBusinessDay` offset with a `BelgianHolidayCalendar` subclass of `AbstractHolidayCalendar`, defining holidays like Easter Monday and Ascension via `Easter` and `Day` offsets. Business days are generated with `pd.date_range`, then resampled monthly with `.resample('M').count()` to produce the counts. The visualization’s data originates from the calendar logic, not an external dataset.

Jan 12, 2017

Loading thumbnail…

fresh block

A single large rectangle sits against a plain background, initially drawn in dark red before transitioning to a bright blue over three seconds. The animation uses a D3 transition with a bounce easing, applied to the rectangle’s fill style. The visualization is built with D3.js (version 3.5.5) and rendered as an SVG element that fills the viewport. The code originates from a Blockbuilder.org gist and demonstrates a minimal, animated color change without any data or interactivity.

May 16, 2016

Loading thumbnail…

Gist feea61c798fdb955f16b5ba785974e1a

This example highlights differences between two pandas DataFrames by key. It displays the two records side-by-side, transposing them into a single table where each column represents a field. Any column containing differing values between the two rows is visually marked with a yellow background. The implementation uses the pandas `DataFrame.style.apply` method to apply a custom `highlight_diff` function that checks for uniqueness across the row pairs, returning conditional formatting styles for the table.

May 10, 2016

Loading thumbnail…

Gist f052b3898971d32fb695

This example uses Python’s matplotlib within a Jupyter notebook to create subplots, leveraging NumPy and matplotlib’s interactive mode. The visualization demonstrates the construction of multiple panels in a single figure, with code likely generating sample plots for comparison or exploration. The notebook runs `%pylab inline` to enable inline plotting and produce rendered subplot images as output.

Feb 26, 2016

Loading thumbnail…

Gist babc8f75061b530146ff

This example demonstrates how a small graph can be converted into a linear system of equations and solved to compute node positions, here yielding values scaled to 50, 100, and 50 for nodes a, b, and c. It shows the process from defining nodes and weighted edges through constructing a matrix equation, solving it with NumPy’s `np.linalg.solve`, and rescaling the result. The visualization relies on Python code and array operations rather than a plotting library, with the final output presented as a list of node–value pairs.

Feb 4, 2016

Loading thumbnail…

Gist 1ba2e157aea3eefadc42

This example shows how a large synthetic file hierarchy is created on a local file system, visualized as a tree-like directory structure. The code uses Python’s `os` module to programmatically create 50 top-level directories, each containing 1,000 subdirectories, and each subdirectory containing 1,000 text files, for a total of 50 million files. The file contents are identical placeholder text, and the visualization likely represents the nested structure using a tree or treemap layout, though the code itself only generates the data, not the graphic. No external visualization libraries are used in the provided script.

Jan 10, 2016

Loading thumbnail…

Alfresco QNames

This visualization maps Alfresco QName relationships as a force-directed graph. It shows how QName entries connect to their namespaces via URI links, with nodes and edges positioned using D3’s `d3.forceSimulation`. The underlying SQL query (`alf_qnames.sql`) joins `alf_qname` to `alf_namespace` to supply the relational data, which is then rendered as an interactive network. Node positions update on each tick of the simulation, while link paths and node radii are drawn using D3’s SVG shape helpers, such as `d3.line` or `d3.arc`, based on the computed force layout.

Oct 26, 2015