#745#775
Conversation
Adds the clrd2025 branch in load_sample mirroring the existing clrd config, but using the modernized CAS Schedule P column names (IncurredLosses rather than IncurLoss). Updates the docstring's complete dataset list and the sample-data documentation page. Adds a targeted test asserting the six LOBs, modern column names, and origin starting at 1998. The underlying clrd2025.csv was added by @kennethshsu on branch #745. Closes part of #745.
feat(data): wire clrd2025 into load_sample, docs, and tests (#745)
kennethshsu
requested review from
genedan,
henrydingliu and
jbogaardt
as code owners
henrydingliu
left a comment
henrydingliu
left a comment
There was a problem hiding this comment.
manifest not updated
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #775 +/- ##
==========================================
+ Coverage 85.94% 86.08% +0.13%
==========================================
Files 85 86 +1
Lines 4924 4923 -1
Branches 637 638 +1
==========================================
+ Hits 4232 4238 +6
+ Misses 491 486 -5
+ Partials 201 199 -2
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
I just tried to play around with it. But what I got (AI assisted) is super messy. What idea do you have to do this cleanly? |
|
i'm good with just adding all the new datasets to manifest.in for this pr. this never went away |
|
Are you still having trouble loading the sample data? It works for me cl.load_sample("friedland_us_industry_auto") |
|
ran this just now. different error though 
|
|
actually, the data is coming with the package. i'm good with this pr 
|
|
I am going to take a wild wild shot here... How did you launch your Jupyter book? Are you doing it via Try |
|
i'm in databricks. i launch jupyter through their UI. the environment is super sterile. all the user installed packages are wiped after each session. here is what it looks like if i just open up a new notebook. it's basically the same as someone who just installed python and installed chainladder for the first time. |
3 participants
Summary of Changes
Added the new CLRD datadump
Related GitHub Issue(s)
Closes #745
Additional Context for Reviewers
uv run pytest) and documentation changes (uv run jb build docs --builder=custom --custom-builder=doctest)Note
Low Risk
Low risk: primarily adds a new packaged CSV sample dataset and small
load_samplerouting/tests/docs updates, with minimal impact beyond sample-data loading.Overview
Adds the new
clrd2025CAS Schedule P refresh sample dataset to the packaged distribution (viaMANIFEST.in) so it ships with the library.Updates
load_sampleto recognizeclrd2025and map it to the refreshed column naming, and adds tests/documentation coverage to ensure the new sample loads and matches expected LOBs/years/columns.Reviewed by Cursor Bugbot for commit f8dd20b. Bugbot is set up for automated code reviews on this repo. Configure here.