Meta Information

How the Data is Prepared

Provenance of the original EVVet data, including the contents of each cell, is considered sacred and treated as immutable. Our end-to-end process involves minimal merging, sanitizing, and augmenting of the data to facilitate analysis as follows:

We did a one-time download of the VeDDRA xlsx file linked on page 2 here, and converted it to a csv while filtering out human terms. The resulting data/output/veddra.csv file is used later; you can see the trivial notebook that does that work here: data/veddra.ipynb.
Next we download csv files from EVVet’s Line Listing Tool. This results in multiple csv files (split by year) that you can find in data/input/evvet/.
The data/fetch-evvet-data.ipynb python notebook then iterates through all the source csv files, sanitizes them (concatenate, rearrange columns, drop duplicates, etc.), generates the data/output/evvet_master.csv output file, and spits out some validation stats to ensure parity with EVVet (which is the source of truth).
evvet_master.csv is then loaded via a CORS proxy into a data variable in this notebook.
A combination of JavaScript, HTML, CSS, Markdown, Observable Plots, and APIs like D3 are used to transform the data as necessary to feed the lazy-loaded, static content blocks in this notebook.

The first three steps are repeated as and when EVVet pushes out more data.

Note On occasion EVVet’s dashboard reflects incorrect metrics as it seems to have a bug whereby the graphs lag a full data update, i.e. the graphs reflect the data from their prior update. We are working on filing a bug against their system to rectify this.

🐙 Fork on Github
The full repository including the source files, fetch/merge code, etc. is available publicly on Github. Feel free to peruse, clone, fork, etc: https://github.com/d8abased/librela

About Us

These notebooks are an ongoing project by volunteer data & computer scientists, with contributions from graphic designers, editors, healthcare professionals, veterinary professionals, and even a geographer. The authors and contributors all have dogs. Some of these dogs have taken Librela, while others’ haven’t; some of these dogs have fared well on Librela, while others have experienced adverse reactions. While the members of this global, interdisciplinary team come from different walks of life, they all share a penchant for data and details.

And of course, dogs. 🐶

📮 Questions or Concerns?
Feel free to contact us at d8abased@gmail.com.