Ace3Z commited on
Commit
0bbccd6
·
verified ·
1 Parent(s): f566690

docs: drop em dashes from the markdown

Browse files

One of them was quoted program output, so rebase_graph_paths.py prints the same wording and the documented output still matches what you see.

Files changed (3) hide show
  1. DATA.md +4 -4
  2. README.md +2 -2
  3. modeling/rebase_graph_paths.py +1 -1
DATA.md CHANGED
@@ -40,13 +40,13 @@ idempotent, so a second run inspects every row and reports the true state:
40
  # second run, complete data
41
  rows inspected : 71745
42
  resolvable now : 71745/71745
43
- all paths already in absolute form and resolvable — nothing to do
44
  ```
45
 
46
  Anything below `N/N` on that second run means graphs are missing. Two traps make `--check`
47
  the wrong tool here:
48
 
49
- - **It stops at 20,000 rows**, and the CSVs are in shard order — the first 20,000 rows of
50
  SPIDER-breast are exactly shards 0-4 of 18. Extract five tars and `--check` prints a clean
51
  `20000/20000` while 51,745 graphs (72 %) are absent. The same applies to BRACS and TCGA-BRCA.
52
  - **It resolves relative paths against the CSV's own directory**, whereas the loader resolves
@@ -274,7 +274,7 @@ cells, columns `centroid_x_px,centroid_y_px,cell_type,type_prob`) plus a summary
274
  `tail_metrics/<cohort>.json`, which is the artifact behind the table above: `nodes` is
275
  `n_nodes_total - n_isolated_total` and the two tail columns are the per-node fractions below
276
  0.3 and 0.5. Do not read `bach_cellvit/bach_fattail_cellvit.json` as the BACH row of that
277
- table — it is a different analysis (whole-image CellViT graphs, 538,421 cells) rather than the
278
  224 px tile graphs (439,533 nodes) the table is computed over.
279
 
280
  The pseudo-labeller was validated on a held-out ground-truth split first: **ARI 0.560**
@@ -310,7 +310,7 @@ Contents: `graphs_v1/` holds NuCLS (1,694 `.pt` × 2 label granularities plus 5-
310
  tables) and PanNuke (3 folds × 20×/40×, 14,414 `.pt`), both with 1-dim edges, and
311
  `source_optuna_hps/` holds the per-task tuned hyperparameters (the search was unseeded, so
312
  these values cannot be recovered by re-running it). Five are Optuna records with the full
313
- trial list; the sixth, `nucls_super_source_args.json`, is not — it is the argument set and
314
  metrics of a single fold-1 run, so for that column the tuned values are recorded but the
315
  search that produced them is not.
316
 
 
40
  # second run, complete data
41
  rows inspected : 71745
42
  resolvable now : 71745/71745
43
+ all paths already in absolute form and resolvable, nothing to do
44
  ```
45
 
46
  Anything below `N/N` on that second run means graphs are missing. Two traps make `--check`
47
  the wrong tool here:
48
 
49
+ - **It stops at 20,000 rows**, and the CSVs are in shard order: the first 20,000 rows of
50
  SPIDER-breast are exactly shards 0-4 of 18. Extract five tars and `--check` prints a clean
51
  `20000/20000` while 51,745 graphs (72 %) are absent. The same applies to BRACS and TCGA-BRCA.
52
  - **It resolves relative paths against the CSV's own directory**, whereas the loader resolves
 
274
  `tail_metrics/<cohort>.json`, which is the artifact behind the table above: `nodes` is
275
  `n_nodes_total - n_isolated_total` and the two tail columns are the per-node fractions below
276
  0.3 and 0.5. Do not read `bach_cellvit/bach_fattail_cellvit.json` as the BACH row of that
277
+ table; it is a different analysis (whole-image CellViT graphs, 538,421 cells) rather than the
278
  224 px tile graphs (439,533 nodes) the table is computed over.
279
 
280
  The pseudo-labeller was validated on a held-out ground-truth split first: **ARI 0.560**
 
310
  tables) and PanNuke (3 folds × 20×/40×, 14,414 `.pt`), both with 1-dim edges, and
311
  `source_optuna_hps/` holds the per-task tuned hyperparameters (the search was unseeded, so
312
  these values cannot be recovered by re-running it). Five are Optuna records with the full
313
+ trial list; the sixth, `nucls_super_source_args.json`, is not: it is the argument set and
314
  metrics of a single fold-1 run, so for that column the tuned values are recorded but the
315
  search that produced them is not.
316
 
README.md CHANGED
@@ -22,7 +22,7 @@ the representation collapsing.
22
 
23
  - **Original paper:** [arXiv:2603.00143](https://arxiv.org/abs/2603.00143) · [model](https://huggingface.co/ogutsevda/graphist)
24
  - **Code:** [Ace3Z/GrapHist-V2](https://github.com/Ace3Z/GrapHist-V2)
25
- - **Data manual:** [`DATA.md`](DATA.md) — every artifact here and the result it reproduces
26
 
27
  ## Results
28
 
@@ -132,7 +132,7 @@ transforms = [
132
  raises no error; it just yields different numbers.
133
 
134
  Three arguments decide whether the weights load at all. `encoder_norm` must be `"layer"` and
135
- `edge_distance_in_proj` must be `False` — both defaults are wrong for this checkpoint — and
136
  `concat_hidden` must be `True`, which has no default. Anything else raises a shape or key error.
137
 
138
  ## Licence and citation
 
22
 
23
  - **Original paper:** [arXiv:2603.00143](https://arxiv.org/abs/2603.00143) · [model](https://huggingface.co/ogutsevda/graphist)
24
  - **Code:** [Ace3Z/GrapHist-V2](https://github.com/Ace3Z/GrapHist-V2)
25
+ - **Data manual:** [`DATA.md`](DATA.md), mapping every artifact here to the result it reproduces
26
 
27
  ## Results
28
 
 
132
  raises no error; it just yields different numbers.
133
 
134
  Three arguments decide whether the weights load at all. `encoder_norm` must be `"layer"` and
135
+ `edge_distance_in_proj` must be `False` (both defaults are wrong for this checkpoint), and
136
  `concat_hidden` must be `True`, which has no default. Anything else raises a shape or key error.
137
 
138
  ## Licence and citation
modeling/rebase_graph_paths.py CHANGED
@@ -103,7 +103,7 @@ def main():
103
  print("check-only; nothing written")
104
  return
105
  if already == n and n:
106
- print(f"all paths already in {args.mode} form and resolvable — nothing to do")
107
  return
108
 
109
  # --- rewrite ---
 
103
  print("check-only; nothing written")
104
  return
105
  if already == n and n:
106
+ print(f"all paths already in {args.mode} form and resolvable, nothing to do")
107
  return
108
 
109
  # --- rewrite ---