add scripts/ctk_structure_viewer.py

fix links to data files in site/src/routes/contribute/+page.md

data/mp/build_phase_diagram.py

@@ -102,7 +102,7 @@
     json.dump(elemental_ref_entries, file, default=lambda x: x.as_dict())
 
 
-df_mp = pd.read_json(DATA_FILES.mp_energies).set_index("material_id")
+df_mp = pd.read_csv(DATA_FILES.mp_energies).set_index("material_id")
 
 
 # %%

data/wbm/analysis.py

@@ -69,7 +69,7 @@
 
 
 # %% load MP training set
-df = pd.read_json(DATA_FILES.mp_energies)
+df = pd.read_csv(DATA_FILES.mp_energies)
 mp_elem_counts = count_elements(df.formula_pretty).astype(int)
 
 # mp_elem_counts.to_json(f"{about_data_page}/mp-element-counts.json")

data/wbm/readme.md

@@ -90,7 +90,7 @@ Element counts for MP training set consisting of 146,323 `ComputedStructureEntri
 
 ## 🎯   Target Distribution
 
-The WBM test set has an energy above the MP convex hull distribution with mean **0.02 eV/atom** and standard deviation of **0.25 eV/atom**.
+The WBM test set has an energy above the MP convex hull distribution with **mean ± std = 0.02 ± 0.25 eV/atom**.
 
 The dummy MAE of always predicting the test set mean is **0.17 eV/atom**.
 

matbench_discovery/preds.py

@@ -33,15 +33,15 @@ class PredFiles(Files):
     bowsr_megnet = "bowsr/2023-01-23-bowsr-megnet-wbm-IS2RE.csv"
     # default CHGNet model from publication with 400,438 params
     chgnet = "chgnet/2023-03-06-chgnet-wbm-IS2RE.csv"
-    # chgnet_megnet = "chgnet/2023-03-04-chgnet-wbm-IS2RE.csv"
+    chgnet_megnet = "chgnet/2023-03-04-chgnet-wbm-IS2RE.csv"
     # CGCnn 10-member ensemble
     cgcnn = "cgcnn/2023-01-26-test-cgcnn-wbm-IS2RE/cgcnn-ensemble-preds.csv"
     # cgcnn 10-member ensemble with 5-fold training set perturbations
     cgcnn_p = "cgcnn/2023-02-05-cgcnn-perturb=5.csv"
     # original m3gnet straight from publication, not re-trained
     m3gnet = "m3gnet/2022-10-31-m3gnet-wbm-IS2RE.csv"
     # m3gnet-relaxed structures fed into megnet for formation energy prediction
-    # m3gnet_megnet = "m3gnet/2022-10-31-m3gnet-wbm-IS2RE.csv"
+    m3gnet_megnet = "m3gnet/2022-10-31-m3gnet-wbm-IS2RE.csv"
     # original megnet straight from publication, not re-trained
     megnet = "megnet/2022-11-18-megnet-wbm-IS2RE/megnet-e-form-preds.csv"
     # magpie composition+voronoi tessellation structure features + sklearn random forest
@@ -123,7 +123,7 @@ def load_df_wbm_preds(
     )
 
 # pick F1 as primary metric to sort by
-df_metrics = df_metrics.round(3).sort_values("F1", axis=1)
+df_metrics = df_metrics.round(3).sort_values("F1", axis=1, ascending=False)
 
 # dataframe of all models' energy above convex hull (EACH) predictions (eV/atom)
 df_each_pred = pd.DataFrame()

matbench_discovery/structure.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import numpy as np
 from pymatgen.core import Structure
 

models/cgcnn/test_cgcnn.py

@@ -49,7 +49,7 @@
     "IS2RE": DATA_FILES.wbm_initial_structures,
     "RS2RE": DATA_FILES.wbm_computed_structure_entries,
     "IS2RE-debug": f"{ROOT}/data/wbm/2022-10-19-wbm-init-structs.json-1k-samples.bz2",
-}[task_type + "-debug" if DEBUG else ""]
+}[task_type + ("-debug" if DEBUG else "")]
 input_col = {"IS2RE": "initial_structure", "RS2RE": "relaxed_structure"}[task_type]
 
 df = pd.read_json(data_path).set_index("material_id")
@@ -69,9 +69,10 @@
     "created_at": {"$gt": "2023-01-09", "$lt": "2023-01-10"},
 }
 runs = wandb.Api().runs(WANDB_PATH, filters=filters)
+expected_runs = 10
 assert (
-    len(runs) == 10
-), f"Expected 10 runs, got {len(runs)} filtering {WANDB_PATH=} with {filters=}"
+    len(runs) == expected_runs
+), f"{expected_runs=}, got {len(runs)} filtering {WANDB_PATH=} with {filters=}"
 
 for idx, run in enumerate(runs):
     for key, val in run.config.items():

models/m3gnet/wbm_pre_vs_post_m3gnet_relaxation.py → models/m3gnet/pre_vs_post_m3gnet_relaxation.py

@@ -47,7 +47,6 @@
     df_m3gnet_is2re.initial_structure.map(lambda x: x["lattice"])
 ).add_prefix("m3gnet_")
 df_m3gnet_is2re[df_m3gnet_lattice.columns] = df_m3gnet_lattice.to_numpy()
-df_m3gnet_is2re
 
 # df_m3gnet_is2re["m3gnet_energy"] = df_m3gnet_is2re.trajectory.map(
 #     lambda x: x["energies"][-1][0]

models/voronoi/train_test_voronoi_rf.py

@@ -48,7 +48,7 @@
 df_train = glob_to_df(train_path).set_index("material_id")
 print(f"{df_train.shape=}")
 
-df_mp = pd.read_json(DATA_FILES.mp_energies).set_index("material_id")
+df_mp = pd.read_csv(DATA_FILES.mp_energies).set_index("material_id")
 train_e_form_col = "formation_energy_per_atom"
 
 test_path = f"{module_dir}/2022-11-18-features-wbm-{task_type}.csv.bz2"

models/wrenformer/test_wrenformer.py

@@ -58,9 +58,10 @@
     "display_name": {"$regex": "wrenformer-robust"},
 }
 runs = wandb.Api().runs(WANDB_PATH, filters=filters)
+expected_runs = 10
 assert (
-    len(runs) == 10
-), f"Expected 10 runs, got {len(runs)} filtering {WANDB_PATH=} with {filters=}"
+    len(runs) == expected_runs
+), f"{expected_runs=}, got {len(runs)} filtering {WANDB_PATH=} with {filters=}"
 
 for idx, run in enumerate(runs):
     for key, val in run.config.items():

scripts/compile_metrics.py

@@ -35,7 +35,7 @@
         ),
     ),
     "CHGNet": dict(
-        n_runs=102,
+        n_runs=100,
         filters=dict(
             display_name={"$regex": "chgnet-wbm-IS2RE-"},
             created_at={"$gt": "2023-03-05", "$lt": "2023-03-07"},

scripts/ctk_structure_viewer.py

@@ -0,0 +1,72 @@
+from __future__ import annotations
+
+import pandas as pd
+from crystal_toolkit.helpers.utils import hook_up_fig_with_struct_viewer
+
+__author__ = "Janosh Riebesell"
+__date__ = "2023-03-07"
+
+"""
+This scripts runs a Crystal Toolkit app that shows a scatter plot of CHGNet energies
+and allows to click on points to view the corresponding structures. Run with:
+python scripts/ctk_structure_viewer.py
+Then open http://localhost:8000 in your browser.
+"""
+
+df_plot = None
+min_e_diff = 0.1
+e_form_2000 = "e_form_per_atom_chgnet_2000"
+e_form_500 = "e_form_per_atom_chgnet_500"
+
+if df_plot is None:
+    from matbench_discovery.preds import PRED_FILES
+
+    df_chgnet = pd.read_json(PRED_FILES.__dict__["CHGNet"].replace(".csv", ".json.gz"))
+    df_chgnet = df_chgnet.set_index("material_id")
+
+    df_chgnet_2000 = pd.read_csv(PRED_FILES.__dict__["CHGNet"])
+    df_chgnet_2000 = df_chgnet_2000.set_index("material_id").add_suffix("_2000")
+    df_chgnet[list(df_chgnet_2000)] = df_chgnet_2000
+
+    df_chgnet_500 = pd.read_csv(PRED_FILES.__dict__["CHGNet"].replace("-06", "-04"))
+    df_chgnet_500 = df_chgnet_500.set_index("material_id").add_suffix("_500")
+    df_chgnet[list(df_chgnet_500)] = df_chgnet_500
+
+    e_form_abs_diff = "e_form_abs_diff"
+    df_chgnet[e_form_abs_diff] = abs(df_chgnet[e_form_2000] - df_chgnet[e_form_500])
+    df_plot = df_chgnet.round(3).query(f"{e_form_abs_diff} > {min_e_diff}")
+
+
+plot_labels = {
+    e_form_500: "CHGNet E<sub>form</sub> after 500 steps",
+    e_form_2000: "CHGNet E<sub>form</sub> after 2000 steps",
+    e_form_abs_diff: "Δ E<sub>form</sub>",
+}
+
+fig = df_plot.reset_index().plot.scatter(
+    x=e_form_500,
+    y=e_form_2000,
+    backend="plotly",
+    hover_name="material_id",
+    hover_data=["formula"],
+    labels=plot_labels,
+    size=e_form_abs_diff,
+    color=e_form_abs_diff,
+    template="plotly_white",
+)
+
+fig.layout.margin.update(b=20, l=40, r=20, t=50)
+fig.layout.coloraxis.colorbar.update(
+    title=dict(text="Energy Diff (eV/atom)", side="right"), thickness=10
+)
+# slightly increase scatter point size (lower sizeref means larger)
+fig.update_traces(marker_sizeref=0.02, selector=dict(mode="markers"))
+
+app = hook_up_fig_with_struct_viewer(
+    fig,
+    df_plot,
+    "chgnet_structure",
+    # validate_id requires material_id to be hover_name
+    validate_id=lambda id: id.startswith(("wbm-", "mp-", "mvc-")),
+)
+app.run_server(debug=True, port=8000)

scripts/difficult_structures.py

@@ -32,7 +32,7 @@
 # %%
 n_rows, n_cols = 5, 4
 for which in ("best", "worst"):
-    fig, axs = plt.subplots(n_rows, n_cols, figsize=(3 * n_rows, 4 * n_cols))
+    fig, axs = plt.subplots(n_rows, n_cols, figsize=(4 * n_cols, 3 * n_rows))
     n_axs = len(axs.flat)
 
     errs = (

scripts/rolling_mae_vs_hull_dist.py

@@ -11,10 +11,11 @@
 
 
 # %%
-# model = "Wrenformer"
+model = "Wrenformer"
 model = "M3GNet + MEGNet"
 model = "MEGNet"
-model = "MEGNet Old"
+model = "CHGNet"
+
 ax, df_err, df_std = rolling_mae_vs_hull_dist(
     e_above_hull_true=df_wbm[each_true_col],
     e_above_hull_errors={model: df_wbm[e_form_col] - df_wbm[model]},

site/src/app.css

@@ -199,8 +199,9 @@ input {
   font-size: 12pt;
   background: rgba(255, 255, 255, 0.1);
   color: var(--text-color);
+  transition: background 0.2s;
 }
 input:focus {
   outline: none;
-  background: rgba(255, 255, 255, 0.15);
+  background: rgba(255, 255, 255, 0.2);
 }

site/src/lib/index.ts

@@ -77,7 +77,8 @@ export type Citation = {
   title: string
   subtitle?: string
   authors: {
-    name: string
+    'family-names': string
+    'given-names': string
     affiliation: string
     affil_key: string
     orcid: string

site/src/routes/contribute/+page.md

@@ -75,13 +75,13 @@ assert list(df_wbm) == [
 
 You can also download the data files directly from GitHub:
 
-1. [`2022-10-19-wbm-summary.csv`]({repo}/raw/v1.0.0/data/wbm/2022-10-19-wbm-summary.csv) [[GitHub]({repo}/blob/v1/data/wbm/2022-10-19-wbm-summary.csv)]: Computed material properties only, no structures. Available properties are VASP energy, formation energy, energy above the convex hull, volume, band gap, number of sites per unit cell, and more. e_form_per_atom and e_above_hull each have 3 separate columns for old, new and no Materials
-1. [`2022-10-19-wbm-init-structs.json`]({repo}/raw/v1.0.0/data/wbm/2022-10-19-wbm-init-structs.json) [[GitHub]({repo}/blob/v1/data/wbm/2022-10-19-wbm-init-structs.json)]: Unrelaxed WBM structures
-1. [`2022-10-19-wbm-cses.json`]({repo}/raw/v1.0.0/data/wbm/2022-10-19-wbm-cses.json) [[GitHub]({repo}/blob/v1/data/wbm/2022-10-19-wbm-cses.json)]: Relaxed WBM structures along with final VASP energies
-1. [`2023-01-10-mp-energies.json.gz`]({repo}/raw/v1.0.0/data/wbm/2023-01-10-mp-energies.json.gz) [[GitHub]({repo}/blob/v1/data/wbm/2023-01-10-mp-energies.json.gz)]: Materials Project formation energies and energies above convex hull
-1. [`2023-02-07-mp-computed-structure-entries.json.gz`]({repo}/raw/v1.0.0/data/wbm/2023-02-07-mp-computed-structure-entries.json.gz) [[GitHub]({repo}/blob/v1/data/wbm/2023-02-07-mp-computed-structure-entries.json.gz)]: Materials Project computed structure entries
-1. [`2023-02-07-ppd-mp.pkl.gz`]({repo}/raw/v1.0.0/data/wbm/2023-02-07-ppd-mp.pkl.gz) [[GitHub]({repo}/blob/v1/data/wbm/2023-02-07-ppd-mp.pkl.gz)]: [PatchedPhaseDiagram](https://pymatgen.org/pymatgen.analysis.phase_diagram.html#pymatgen.analysis.phase_diagram.PatchedPhaseDiagram) constructed from all MP ComputedStructureEntries
-1. [`2022-09-19-mp-elemental-ref-energies.json`]({repo}/raw/v1.0.0/data/wbm/2022-09-19-mp-elemental-ref-energies.json) [[GitHub]({repo}/blob/v1/data/wbm/2022-09-19-mp-elemental-ref-energies.json)]: Minimum energy PDEntries for each element present in the Materials Project
+1. [`2022-10-19-wbm-summary.csv`]({repo}/blob/-/data/wbm/2022-10-19-wbm-summary.csv): Computed material properties only, no structures. Available properties are VASP energy, formation energy, energy above the convex hull, volume, band gap, number of sites per unit cell, and more. e_form_per_atom and e_above_hull each have 3 separate columns for old, new and no Materials
+1. [`2022-10-19-wbm-init-structs.json`]({repo}/blob/-/data/wbm/2022-10-19-wbm-init-structs.json): Unrelaxed WBM structures
+1. [`2022-10-19-wbm-cses.json`]({repo}/blob/-/data/wbm/2022-10-19-wbm-cses.json): Relaxed WBM structures along with final VASP energies
+1. [`2023-01-10-mp-energies.json.gz`]({repo}/blob/-/data/mp/2023-01-10-mp-energies.json.gz): Materials Project formation energies and energies above convex hull
+1. [`2023-02-07-mp-computed-structure-entries.json.gz`]({repo}/blob/-/data/mp/2023-02-07-mp-computed-structure-entries.json.gz): Materials Project computed structure entries
+1. [`2023-02-07-ppd-mp.pkl.gz`]({repo}/blob/-/data/mp/2023-02-07-ppd-mp.pkl.gz): [PatchedPhaseDiagram](https://pymatgen.org/pymatgen.analysis.phase_diagram.html#pymatgen.analysis.phase_diagram.PatchedPhaseDiagram) constructed from all MP ComputedStructureEntries
+1. [`2022-09-19-mp-elemental-reference-entries.json`]({repo}/blob/-/data/mp/2022-09-19-mp-elemental-reference-entries.json): Minimum energy PDEntries for each element present in the Materials Project
 
 [wbm paper]: https://nature.com/articles/s41524-020-00481-6
 

site/src/routes/models/+page.server.ts

@@ -1,10 +1,9 @@
 import type { ModelData } from '$lib'
 import { compile } from 'mdsvex'
 import { dirname } from 'path'
-import type { PageServerLoad } from './$types'
 import model_stats from './model-stats.json'
 
-export const load: PageServerLoad = async () => {
+export const load = async () => {
   const yml = import.meta.glob(`$root/models/**/metadata.yml`, {
     eager: true,
   })

site/src/routes/models/+page.svelte

@@ -5,9 +5,9 @@
   import { RadioButtons, Tooltip } from 'svelte-zoo'
   import { flip } from 'svelte/animate'
   import { fade } from 'svelte/transition'
-  import type { PageData, Snapshot } from './$types'
+  import type { Snapshot } from './$types'
 
-  export let data: PageData
+  export let data
 
   let sort_by: keyof ModelStats | 'model_name' = `F1`
   let show_details = false
@@ -125,4 +125,10 @@
   span :global(div.zoo-radio-btn span) {
     padding: 1pt 4pt;
   }
+  input[type='number'] {
+    text-align: center;
+  }
+  input[type='number']::-webkit-inner-spin-button {
+    display: none;
+  }
 </style>

site/src/routes/paper/+layout.server.ts

@@ -1,13 +1,13 @@
+import type { Citation } from '$lib'
 import fs from 'fs'
 import yml from 'js-yaml'
-import type { LayoutServerLoad } from './$types'
 
-export const load: LayoutServerLoad = async ({ route }) => {
+export const load = async ({ route }) => {
   const data = fs.readFileSync(`src/routes/${route.id}/+page.md`, `utf8`)
   const cff = fs.readFileSync(`../citation.cff`, `utf8`)
 
   // Count the number of words using a regular expression
   const word_count = data.match(/\b\w+\b/g)?.length ?? null
 
-  return { word_count, ...yml.load(cff) }
+  return { word_count, ...(yml.load(cff) as Citation) }
 }

site/src/routes/paper/+layout.svelte

@@ -1,10 +1,9 @@
 <script lang="ts">
   import { References } from '$lib'
   import { pretty_num } from 'elementari/labels'
-  import type { LayoutServerData } from './$types'
   import { references } from './references.yaml'
 
-  export let data: LayoutServerData
+  export let data
 
   const authors = data.authors.map(
     (auth) => `${auth[`given-names`]} ${auth[`family-names`]}<sup>${auth.affil_key}</sup>`