Problem
Public docstring coverage is 239 of 297 public non-dunder definitions (80.5%). The 58 gaps concentrate in two modules:
| Module |
Missing |
_graph_tools.py |
17 (mostly properties) |
markov/core.py |
16 |
_arma.py, _kalman.py, distributions.py, util/timing.py |
4 each |
markov/ddp.py, random/utilities.py, others |
3 each |
Where docstrings exist they are numpydoc-conformant and generally good; this is about completeness, not quality.
Good for outside contributors, especially anyone with domain knowledge of the graph and Markov chain code, since several gaps are on properties whose meaning is not obvious from the name. Worth splitting into per-module sub-issues so two people do not collide.
Acceptance criteria
From the July 2026 technical-debt audit (AI-assisted; claims verified against 28d4b3b on 2026-07-25).
Problem
Public docstring coverage is 239 of 297 public non-dunder definitions (80.5%). The 58 gaps concentrate in two modules:
_graph_tools.pymarkov/core.py_arma.py,_kalman.py,distributions.py,util/timing.pymarkov/ddp.py,random/utilities.py, othersWhere docstrings exist they are numpydoc-conformant and generally good; this is about completeness, not quality.
Good for outside contributors, especially anyone with domain knowledge of the graph and Markov chain code, since several gaps are on properties whose meaning is not obvious from the name. Worth splitting into per-module sub-issues so two people do not collide.
Acceptance criteria
From the July 2026 technical-debt audit (AI-assisted; claims verified against
28d4b3bon 2026-07-25).