v1.1.2

Author: janosh

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ default_install_hook_types: [pre-commit, commit-msg]
 
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.4.4
+    rev: v0.4.6
     hooks:
       - id: ruff
         args: [--fix]
@@ -37,7 +37,7 @@ repos:
         additional_dependencies: [types-pyyaml, types-requests]
 
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.2.6
+    rev: v2.3.0
     hooks:
       - id: codespell
         stages: [commit, commit-msg]
@@ -57,7 +57,7 @@ repos:
         exclude: ^(site/src/figs/.+\.svelte|data/wbm/20.+\..+|site/src/(routes|figs).+\.(yaml|json)|changelog.md)$
 
   - repo: https://github.com/pre-commit/mirrors-eslint
-    rev: v9.2.0
+    rev: v9.3.0
     hooks:
       - id: eslint
         types: [file]
@@ -71,15 +71,15 @@ repos:
           - typescript-eslint
 
   - repo: https://github.com/python-jsonschema/check-jsonschema
-    rev: 0.28.3
+    rev: 0.28.4
     hooks:
       - id: check-jsonschema
         files: ^models/(.+)/\1.*\.yml$
         args: [--schemafile, tests/model-schema.yml]
       - id: check-github-actions
 
   - repo: https://github.com/RobertCraigie/pyright-python
-    rev: v1.1.363
+    rev: v1.1.365
     hooks:
       - id: pyright
         args: [--level, error]

changelog.md

@@ -2,6 +2,35 @@
 
 All notable changes to this project will be documented in this file. Dates are displayed in UTC.
 
+#### [v1.1.2](https://github.com/janosh/matbench-discovery/compare/v1.1.1...v1.1.2)
+
+> 15 May 2024
+
+- Add `pyright` pre-commit hook and fix possibly unbound variables [`#93`](https://github.com/janosh/matbench-discovery/pull/93)
+- Fix landing page vs model page metrics mismatch [`#92`](https://github.com/janosh/matbench-discovery/pull/92)
+- Fill 3% missing MACE preds [`#90`](https://github.com/janosh/matbench-discovery/pull/90)
+- doc: add prototype description [`#89`](https://github.com/janosh/matbench-discovery/pull/89)
+- Rough solution to recreating plots for unique proto set [`#88`](https://github.com/janosh/matbench-discovery/pull/88)
+- Pass preds rather than err to v-plots [`#87`](https://github.com/janosh/matbench-discovery/pull/87)
+- Extend model metadata schema [`#86`](https://github.com/janosh/matbench-discovery/pull/86)
+- Model metadata schema validation [`#85`](https://github.com/janosh/matbench-discovery/pull/85)
+- Add GNoME model [`#84`](https://github.com/janosh/matbench-discovery/pull/84)
+- Fix test scripts raising `wandb.Table` `AssertionError`: columns argument expects list of strings or ints [`#82`](https://github.com/janosh/matbench-discovery/pull/82)
+- Fix landing page vs model page metrics mismatch (#92) [`#91`](https://github.com/janosh/matbench-discovery/issues/91)
+- fix CaptionedMetricsTable proprietary models toggle [`64e97af`](https://github.com/janosh/matbench-discovery/commit/64e97af41f6e08e4c51cef394fb9c914b0af3830)
+- hide test_set_std_key from <Select /> in Per-Element Model Error Heatmaps on /models page [`07555b5`](https://github.com/janosh/matbench-discovery/commit/07555b54733164090c033bb8c186c7ef1443d605)
+- fix pymatviz bin_df_cols util imports + ruff fixes + bump site deps [`a3b4362`](https://github.com/janosh/matbench-discovery/commit/a3b43624b5bba0562bdbf5b0fb9fc15cefb8ffbe)
+- bump ruff target-version = "py311" [`e8a4444`](https://github.com/janosh/matbench-discovery/commit/e8a4444741da9996fcc2fa83f73b035ad219fab8)
+- add module doc string with installation instructions to train_mace.py [`29eecf2`](https://github.com/janosh/matbench-discovery/commit/29eecf28f756289e4b8a286d00e7c20d219e496a)
+- plot script fixes [`202a305`](https://github.com/janosh/matbench-discovery/commit/202a3058d428dc7867ba6b0105d087cf29c7e130)
+- change TL;DR to MP non-endorsement disclaimer [`016e71f`](https://github.com/janosh/matbench-discovery/commit/016e71fac0040edaad2b73df8aeea61e5097d453)
+- fix and unignore ruff DTZ005 FBT001 FBT002 [`0249327`](https://github.com/janosh/matbench-discovery/commit/0249327b6b5a1eb7578b99ea95613d50d2465392)
+- fix email and MP DB version used for convex hull construction in list of WBM processing steps [`1cc9fe6`](https://github.com/janosh/matbench-discovery/commit/1cc9fe614f431fd9fdae96512dc7bbde5137c81c)
+- fix model stats JSON export in compile_model_stats.py and use model-stats-uniq-protos.json on landing page to match metrics table [`4d40cd6`](https://github.com/janosh/matbench-discovery/commit/4d40cd61eb529d0242756ccb5a0cd87f732e4b59)
+- bump ruff pre-commit to 0.3.0 + auto-fixes [`71a5edc`](https://github.com/janosh/matbench-discovery/commit/71a5edc666a50bcab757bd5211886610677ab24e)
+- fix metrics-table horizontal scrolling [`a5a7bda`](https://github.com/janosh/matbench-discovery/commit/a5a7bdaefc781a14453aa604689e5131425ee67a)
+- fix copy-paste error of trained_by section in gnome.yml [`0ade5d6`](https://github.com/janosh/matbench-discovery/commit/0ade5d65a3544502649b00f756c984125463fe38)
+
 #### [v1.1.1](https://github.com/janosh/matbench-discovery/compare/v1.1.0...v1.1.1)
 
 > 28 January 2024

matbench_discovery/metrics.py

@@ -122,7 +122,7 @@ def stable_metrics(
     if TPR + FNR != 1:  # sanity check: true positives + false negatives = all positives
         raise ValueError(f"{TPR=} {FNR=} don't add up to 1")
 
-    # Drop NaNs to calculat the regression metrics
+    # Drop NaNs to calculate regression metrics
     is_nan = np.isnan(each_true) | np.isnan(each_pred)
     each_true, each_pred = np.array(each_true)[~is_nan], np.array(each_pred)[~is_nan]
 

models/wrenformer/analyze_wrenformer.py

@@ -75,8 +75,9 @@
 styler = df_proto_counts.head(10).style.background_gradient(cmap="viridis")
 styler.set_caption(f"Top 10 {proto_col} in {len(df_bad)} {model} failures")
 display(styler)
-df_to_html_table(styler, f"{SITE_FIGS}/proto-counts-{model_low}-failures.svelte")
-df_to_pdf(styler, f"{PDF_FIGS}/proto-counts-{model_low}-failures.pdf")
+img_name = f"proto-counts-{model_low}-failures"
+df_to_html_table(styler, file_path=f"{SITE_FIGS}/{img_name}.svelte")
+df_to_pdf(styler, f"{PDF_FIGS}/{img_name}.pdf")
 
 
 # %%

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "matbench-discovery"
-version = "1.1.1"
+version = "1.1.2"
 description = "A benchmark for machine learning energy models on inorganic crystal stability prediction from unrelaxed structures"
 authors = [{ name = "Janosh Riebesell", email = "janosh.riebesell@gmail.com" }]
 readme = "readme.md"

scripts/model_figs/metrics_tables.py

@@ -183,7 +183,7 @@
     }"""
     df_to_html_table(
         styler,
-        f"{SITE_FIGS}/metrics-table{label}.svelte",
+        file_path=f"{SITE_FIGS}/metrics-table{label}.svelte",
         inline_props="class='metrics'",
         # draw line between classification and regression metrics
         styles=f"{col_selector} {{ border-left: 1px solid white; }}{hide_scroll_bar}",

scripts/upload_to_figshare.py

@@ -36,7 +36,7 @@ def make_request(
     headers = {"Authorization": f"token {TOKEN}"}
     if data is not None and not binary:
         data = json.dumps(data)
-    response = requests.request(method, url, headers=headers, data=data)
+    response = requests.request(method, url, headers=headers, data=data, timeout=10)
     try:
         response.raise_for_status()
         try:

site/src/routes/preprint/references.yaml

@@ -859,8 +859,7 @@ references:
   source: Cambridge University Press
   title: 'NOMAD: The FAIR concept for big data-driven materials science'
   title-short: NOMAD
-  URL: >-
-    https://www.cambridge.org/core/journals/mrs-bulletin/article/nomad-the-fair-concept-for-big-datadriven-materials-science/1EEF321F62D41997CA16AD367B74C4B0
+  URL: https://doi.org/10.1557/mrs.2018.208
   volume: '43'
 
 - id: dunn_benchmarking_2020
@@ -1184,7 +1183,7 @@ references:
     Commentary: The Materials Project: A materials genome approach to
     accelerating materials innovation
   title-short: Commentary
-  URL: https://aip.scitation.org/doi/10.1063%2F1.4812323
+  URL: https://doi.org/10.1063/1.4812323
   volume: '1'
 
 - id: kovacs_evaluation_2023
@@ -1451,8 +1450,7 @@ references:
   title: >-
     Direction-based Graph Representation to Accelerate Stable Catalysts
     Discovery
-  URL: >-
-    https://chemrxiv.org/engage/chemrxiv/article-details/633660f9114b7ebf1a229015
+  URL: https://doi.org/10.26434/chemrxiv-2022-dp58c
 
 - id: park_developing_2020
   accessed:

tests/test_plots.py

@@ -18,7 +18,7 @@
 
 AxLine = Literal["x", "y", "xy", ""]
 models = ["MEGNet", "CGCNN", "Voronoi RF"]
-df_wbm = load_df_wbm_with_preds(models, nrows=100)
+df_wbm = load_df_wbm_with_preds(models=models, nrows=100)
 
 
 @pytest.mark.parametrize(

tests/test_preds.py

@@ -46,7 +46,7 @@ def test_df_each_err() -> None:
 
 @pytest.mark.parametrize("models", [[], ["Wrenformer"]])
 def test_load_df_wbm_with_preds(models: list[str]) -> None:
-    df = load_df_wbm_with_preds(models)
+    df = load_df_wbm_with_preds(models=models)
     assert len(df) == len(df_wbm)
     assert list(df) == list(df_wbm) + models + [f"{model}_std" for model in models]
     assert df.index.name == Key.mat_id