Documentation a newcomer actually needs
Browse files
README.md
CHANGED
|
@@ -231,6 +231,37 @@ an impossible bound, tampers with a published value, and requires the checker to
|
|
| 231 |
reject both — then confirms an isolated pair reproduces `k = 1.000000000000`
|
| 232 |
exactly, since a lone pair has nothing to screen it.
|
| 233 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 234 |
## Provenance
|
| 235 |
|
| 236 |
Exported by `export.py` from a committed proof witness produced by an
|
|
|
|
| 231 |
reject both — then confirms an isolated pair reproduces `k = 1.000000000000`
|
| 232 |
exactly, since a lone pair has nothing to screen it.
|
| 233 |
|
| 234 |
+
## Troubleshooting
|
| 235 |
+
|
| 236 |
+
**`verify.py` reports violations** — that is the interesting outcome, and we
|
| 237 |
+
want to hear about it. Sampling cannot prove the bound but it can refute it, so
|
| 238 |
+
a genuine violation means the theorem is wrong. Before reporting, re-run with
|
| 239 |
+
`--self-test` to confirm the checker still discriminates: a checker that has
|
| 240 |
+
stopped working can produce either verdict.
|
| 241 |
+
|
| 242 |
+
**`ModuleNotFoundError: numpy` from `verify.py`** — it should never import
|
| 243 |
+
numpy. If it does, the file has been edited; the shipped version runs under
|
| 244 |
+
`env -i /usr/bin/python3` with nothing installed, and a test asserts it imports
|
| 245 |
+
no third-party module.
|
| 246 |
+
|
| 247 |
+
**`loader.to_pandas` or `to_hf_dataset` raises ImportError** — those two are
|
| 248 |
+
conveniences and do need `pandas` / `datasets`. Everything else, including
|
| 249 |
+
`verify.py`, is standard library only.
|
| 250 |
+
|
| 251 |
+
**The Hub viewer shows two configs and you wanted one table** — `regions` and
|
| 252 |
+
`counterexamples` have different schemas and are deliberately separate. Pick the
|
| 253 |
+
config in the viewer's dropdown, or use `load_regions()` / `load_counterexamples()`.
|
| 254 |
+
|
| 255 |
+
**`export.py --check` says the data does not match** — it re-derives the files
|
| 256 |
+
from the committed proof witness and compares. A mismatch means either the data
|
| 257 |
+
or the witness was edited. Counterexample regeneration also needs `maxwell-lint`
|
| 258 |
+
installed, since they are produced by running its reference models.
|
| 259 |
+
|
| 260 |
+
**You want every certified leaf, not the 256 regions** — they are not published.
|
| 261 |
+
The source proof records per-region certification plus aggregate counts, so the
|
| 262 |
+
237,490 leaves are attested but not enumerated here. That is a real limit of
|
| 263 |
+
this release and is stated rather than glossed.
|
| 264 |
+
|
| 265 |
## Provenance
|
| 266 |
|
| 267 |
Exported by `export.py` from a committed proof witness produced by an
|