add ml_stability/m3gnet/plots/2022-08-02-m3gnet-wbm-hull-dist-hist.pdf

and plots/from-rhys/2022-08-06-hist-wbm-batches-true-energy-all-elements.pdf

Author: janosh

ml_stability/__init__.py

@@ -1,8 +1,21 @@
 import os
 from os.path import dirname
+from typing import Any, Generator, Sequence
 
 
 PKG_DIR = dirname(__file__)
 ROOT = dirname(PKG_DIR)
 
 os.makedirs(f"{PKG_DIR}/plots", exist_ok=True)
+
+
+def chunks(xs: Sequence[Any], n: int) -> Generator[Sequence[Any], None, None]:
+    return (xs[i : i + n] for i in range(0, len(xs), n))
+
+
+def as_dict_handler(obj: Any) -> dict[str, Any] | None:
+    """Use as default_handler kwarg to json.dump() or pandas.to_json()."""
+    try:
+        return obj.as_dict()  # all MSONable objects implement as_dict()
+    except AttributeError:
+        return None  # replace unhandled objects with None in serialized data

ml_stability/wbm_pre_vs_post_m3gnet_relaxation.py

@@ -24,34 +24,34 @@
 ).set_index("material_id")
 
 
-# %% spread M3GNet post-pseudo-relaxation lattice params into separate columns
+# %%
 df_m3gnet = pd.read_json(
-    f"{ROOT}/data/2022-07-17@20-27-m3gnet-wbm-relax-results.json.gz"
+    f"{ROOT}/data/2022-07-17-m3gnet-wbm-relax-results.json.gz"
 ).set_index("material_id")
 
 print("Number of WBM crystals for which we have M3GNet results:")
 print(f"{len(df_m3gnet):,} / {len(df_wbm):,} = {len(df_m3gnet)/len(df_wbm):.1%}")
 
 
-# %%
+# %% spread M3GNet post-pseudo-relaxation lattice params into separate columns
 df_m3gnet["final_energy"] = df_m3gnet.trajectory.map(lambda x: x["energies"][-1][0])
 
 df_m3gnet_lattice = pd.json_normalize(
-    df_m3gnet.final_structure.map(lambda x: x["lattice"])
+    df_m3gnet.initial_structure.map(lambda x: x["lattice"])
 ).add_prefix("m3gnet_")
 df_m3gnet[df_m3gnet_lattice.columns] = df_m3gnet_lattice.to_numpy()
 df_m3gnet
 
 
 # %% spread WBM initial and final lattice params into separate columns
-df_m3gnet["cse"] = df_wbm.cse
+df_m3gnet["final_wbm_structure"] = df_wbm.cse.map(lambda x: x["structure"])
 df_wbm_final_lattice = pd.json_normalize(
-    df_m3gnet.cse.map(lambda x: x["structure"]["lattice"])
+    df_m3gnet.final_wbm_structure.map(lambda x: x["lattice"])
 ).add_prefix("final_wbm_")
 df_m3gnet[df_wbm_final_lattice.columns] = df_wbm_final_lattice.to_numpy()
 
 
-df_m3gnet["initial_structure"] = df_wbm.initial_structure
+df_m3gnet["initial_wbm_structure"] = df_wbm.initial_structure
 df_wbm_initial_lattice = pd.json_normalize(
     df_m3gnet.initial_structure.map(lambda x: x["lattice"])
 ).add_prefix("initial_wbm_")
@@ -67,7 +67,7 @@
 df_wbm_initial_lattice = pd.json_normalize(
     df_wbm.initial_structure.map(lambda x: x["lattice"])
 ).add_prefix("initial_wbm_")
-df_wbm[df_wbm_initial_lattice.columns] = df_wbm_initial_lattice
+df_wbm[df_wbm_initial_lattice.columns] = df_wbm_initial_lattice.to_numpy()
 
 print(f"{df_wbm.isna().sum()=}")
 
@@ -102,7 +102,6 @@
 fig = px.histogram(
     df_m3gnet.filter(like="alpha"), nbins=1000, barmode="overlay", log_y=True
 )
-# fig.write_image("plots/alpha-wbm-angles.png", scale=2)
 fig.show()