fix string->int failure in bad id test

[alessandrofelder]

Pandas now enforces "banning of upcasting in in-place setitem-like operations; see PDEP6 (GH 59007)" since release 3.0, which we were doing in this test.

I fixed this by, instead of pretending the string "BAD ID" is an int64, setting the bad id to the most negative value that we can represent with np.int64 (which is indeed a bad id for the Allen morphology neurons).

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 72.37%. Comparing base (18f7178) to head (b587719).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #106   +/-   ##
=======================================
  Coverage   72.37%   72.37%           
=======================================
  Files          10       10           
  Lines         724      724           
=======================================
  Hits          524      524           
  Misses        200      200           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[IgorTatarnikov]

LGTM!