add site/src/figs/(largest-fp-diff-each-error-models|largest-each-errors-fp-diff-models).svelte

shown on /models/tmi page (ex /models/per-element)
generated by scripts/difficult_structures.py

add col site_stats_fingerprint_init_final_norm_diff to data/wbm/2022-10-19-wbm-summary.csv

sort EACH scatter plot facets and legend by MAE

Author: janosh

.gitattributes

@@ -1,3 +1,3 @@
 # exclude generated plot files when calculating repo language statistics on GitHub
-*/figs/* linguist-generated
+**/figs/* linguist-generated
 data/**/*.svelte linguist-generated

data/mp/get_mp_energies.py

@@ -4,7 +4,7 @@
 import pandas as pd
 from aviary.wren.utils import get_aflow_label_from_spglib
 from mp_api.client import MPRester
-from pymatviz.utils import annotate_mae_r2
+from pymatviz.utils import annotate_metrics
 from tqdm import tqdm
 
 from matbench_discovery import today
@@ -71,7 +71,7 @@
     title=f"{today} - {len(df):,} MP entries",
 )
 
-annotate_mae_r2(df.formation_energy_per_atom, df.decomposition_enthalpy)
+annotate_metrics(df.formation_energy_per_atom, df.decomposition_enthalpy)
 # result on 2023-01-10: plots match. no correlation between formation energy and
 # decomposition enthalpy. R^2 = -1.571, MAE = 1.604
 # ax.figure.savefig(f"{module_dir}/mp-decomp-enth-vs-e-form.webp", dpi=300)

data/wbm/fetch_process_wbm_dataset.py

@@ -6,6 +6,7 @@
 import warnings
 from glob import glob
 
+import numpy as np
 import pandas as pd
 from aviary.wren.utils import get_aflow_label_from_spglib
 from pymatgen.analysis.phase_diagram import PatchedPhaseDiagram
@@ -597,6 +598,18 @@ def fix_bad_struct_index_mismatch(material_id: str) -> str:
 assert df_summary[wyckoff_col].isna().sum() == 0
 
 
+# %% site-stats.json.gz was generated by scripts/compute_struct_fingerprints.py
+df_fp = pd.read_json(f"{module_dir}/site-stats.json.gz").set_index("material_id")
+init_fp_col = "initial_site_stats_fingerprint"
+final_fp_col = "final_site_stats_fingerprint"
+fp_diff_col = "site_stats_fingerprint_init_final_norm_diff"
+df_fp[fp_diff_col] = (
+    df_fp[final_fp_col].map(np.array) - df_fp[init_fp_col].map(np.array)
+).map(np.linalg.norm)
+
+df_fp[fp_diff_col].hist(bins=100, backend="plotly")
+
+
 # %% write final summary data to disk (yeah!)
 df_summary.round(6).to_csv(f"{module_dir}/{today}-wbm-summary.csv")
 

scripts/compute_struct_fingerprints.py

@@ -7,6 +7,7 @@
 # %%
 import os
 import warnings
+from glob import glob
 
 import numpy as np
 import pandas as pd
@@ -46,6 +47,7 @@
     account="LEE-SL3-CPU",
     time="6:0:0",
     array=f"1-{slurm_array_task_count}",
+    slurm_flags=("--mem", "30G"),
 )
 
 
@@ -95,4 +97,28 @@
         except Exception as exc:
             print(f"{fp_col} for {row.Index} failed: {exc}")
 
-df_in.filter(like="site_stats_fingerprint").to_json(out_path)
+df_in.filter(like="site_stats_fingerprint").reset_index().to_json(out_path)
+
+
+# %%
+running_as_slurm_job = os.getenv("SLURM_JOB_ID")
+if running_as_slurm_job:
+    print(f"Job wrote {out_path=} and finished at {timestamp}")
+    raise SystemExit(0)
+
+
+# %%
+out_files = glob(f"{out_dir}/site-stats-*.json.gz")
+
+found_idx = [int(name.split("-")[-1].split(".")[0]) for name in out_files]
+print(f"Found {len(out_files)=:,}")
+missing_files = sorted(set(range(1, slurm_array_task_count + 1)) - set(found_idx))
+if missing_files:
+    print(f"{len(missing_files)=}: {missing_files}")
+
+df_out = pd.concat(pd.read_json(out_file) for out_file in tqdm(out_files))
+
+
+df_out.index.name = "material_id"
+
+df_out.reset_index().to_json(f"{out_dir}/site-stats.json.gz")