Added benchmark CRBH20 reaction energy barriers dataset by Quantumplations · Pull Request #343 · ddmms/ml-peg

Quantumplations · 2026-02-04T15:55:53Z

Pre-review checklist for PR author

PR author must check the checkboxes below when creating the PR.

I've confirmed the contribution guidelines.

Summary

Input data: CRBH20 (Chemical Reaction Barrier Heights 20), a set of 20 molecular geometries of reactants and their transition states (cycloreversion of heterocyclic rings).
Calculations: predicted reactant and transition state energies using 6 foundation models (mace-mp-0a, mace-mp-0b3, mace-mpa-0, mace-omat-0, mace-matpes-r2scan, orb-v3-consv-inf-omat, pet-mad), along with the reaction barrier height (difference between transition state and reactant energies).
Analysis: compare the 6 models' predicted reaction barrier heights against r2SCAN DFT reference barrier heights, metric for each model is its MAE over the 20 systems.

Linked issue

Resolves 1 issue.

Resolves #291

Progress

Calculations
Analysis
Application
Documentation

Testing

mace-mp-0a, mace-mp-0b3, mace-mpa-0, mace-omat-0, mace-matpes-r2scan, orb-v3-consv-inf-omat, pet-mad

New decorators/callbacks

…c_rxn_barriers.py script

…script

…rrect display of model MAEs in browser

Quantumplations · 2026-05-14T10:49:47Z

Hey sorry for the long delay in getting on top of this! I've rebased the PR, changed the naming to CRBH20, did the pre-commit formatting, and also I removed some raw data from the repo. I'm also attaching the data zip file for the S3 bucket. Please review and let me know if there's anything else that needs to be done!
CRBH20.zip

joehart2001 · 2026-05-14T13:24:44Z

+    if hasattr(model, "add_d3_calculator"):
+        calc = model.add_d3_calculator(calc)


Suggested change

if hasattr(model, "add_d3_calculator"):

calc = model.add_d3_calculator(calc)

calc = model.add_d3_calculator(calc)

this should be handled

joehart2001 · 2026-05-14T13:26:22Z

+
+# ml_peg imports for model management
+from ml_peg.models.get_models import load_models
+from ml_peg.models.models import current_models


Suggested change

from ml_peg.models.models import current_models

from ml_peg.models import current_models

due to changes we made

joehart2001 · 2026-05-14T13:27:12Z

+from ase.io import read, write
+import pytest
+
+# ml_peg imports for model management


Suggested change

# ml_peg imports for model management

from ml_peg.calcs.utils.utils import download_s3_data

joehart2001 · 2026-05-14T13:27:29Z

+    # Apply D3 dispersion if the model supports/requires it (standard ml-peg pattern)
+    if hasattr(model, "add_d3_calculator"):
+        calc = model.add_d3_calculator(calc)
+


Suggested change

data_dir = (

download_s3_data(

key="inputs/molecular/CRBH20/CRBH20.zip",

filename="CRBH20.zip",

)

/ "CRBH20"

)

joehart2001 · 2026-05-14T13:28:48Z

+    # 2. Loop through Reaction Folders (1 to 20)
+    for i in range(1, 21):
+        rxn_id = str(i)
+        rxn_path = DATA_PATH / rxn_id


Suggested change

rxn_path = DATA_PATH / rxn_id

rxn_path = data_dir / rxn_id

joehart2001 · 2026-05-14T13:29:23Z

+# """Run calculations of reaction energy barriers for 20 systems"""
+
+# from __future__ import annotations
+
+# from copy import copy
+# from pathlib import Path
+# from typing import Any
+
+# from ase import units
+# from ase.io import read, write
+# import numpy as np
+# import pytest
+
+# from ml_peg.calcs.utils.utils import download_s3_data
+# from ml_peg.models.get_models import load_models
+# from ml_peg.models.models import current_models
+
+# MODELS = load_models(current_models)
+
+# DATA_PATH = Path(__file__).parent / "data"
+# OUT_PATH = Path(__file__).parent / "outputs"
+
+# # Unit conversion
+# EV_TO_KJ_PER_MOL = units.mol / units.kJ
+
+# @pytest.mark.parametrize("mlip", MODELS.items())
+# def test_rxn_barrier_calculation(mlip: tuple[str, Any]) -> None:
+#     """
+#     Run calculations of the reaction energy barriers for the 20 systems in CRBH20.
+
+#     Parameters
+#     ----------
+#     mlip
+#         Name of model use and model to get calculator.
+#     """
+#     model_name, model = mlip
+#     calc = model.get_calculator()
+
+#     # Add D3 calculator for this test (for models where applicable)
+#     calc = model.add_d3_calculator(calc)
+
+#     # Download X23 dataset
+#     lattice_energy_dir = (
+#         download_s3_data(
+#             key="inputs/molecular_crystal/X23/X23.zip",
+#             filename="lattice_energy.zip",
+#         )
+#         / "lattice_energy"
+#     )
+
+#     with open(lattice_energy_dir / "list") as f:
+#         systems = f.read().splitlines()
+
+#     for system in systems:
+#         molecule_path = lattice_energy_dir / system / "POSCAR_molecule"
+#         solid_path = lattice_energy_dir / system / "POSCAR_solid"
+#         ref_path = lattice_energy_dir / system / "lattice_energy_DMC"
+#         num_molecules_path = lattice_energy_dir / system / "nmol"
+
+#         molecule = read(molecule_path, index=0, format="vasp")
+#         molecule.calc = calc
+#         molecule.get_potential_energy()
+
+#         solid = read(solid_path, index=0, format="vasp")
+#         solid.calc = copy(calc)
+#         solid.get_potential_energy()
+
+#         ref = np.loadtxt(ref_path)[0]
+#         num_molecules = np.loadtxt(num_molecules_path)
+
+#         solid.info["ref"] = ref
+#         solid.info["num_molecules"] = num_molecules
+#         solid.info["system"] = system
+#         molecule.info["ref"] = ref
+#         molecule.info["num_molecules"] = num_molecules
+#         molecule.info["system"] = system
+
+#         # Write output structures
+#         write_dir = OUT_PATH / model_name
+#         write_dir.mkdir(parents=True, exist_ok=True)
+#         write(write_dir / f"{system}.xyz", [solid, molecule])
+
+
+# import os
+# from mace.calculators import mace_mp
+# from ase.io import read
+
+# # 1. Initialize the model ONCE (much faster than reloading it 40 times)
+# # Define the path to your downloaded model
+# # Make sure the filename matches exactly what you downloaded
+# model_path = os.path.abspath("models/mace-mp-0b3-medium.model")
+
+# print(f"Loading MACE model from: {model_path}")
+
+# # Load the model using the path
+# macemp = mace_mp(
+#     model=model_path,     # <--- Point to the local file here
+#     dispersion=True,      # Keep dispersion on
+#     default_dtype="float64",
+#     device="cuda"         # Use "cpu" if you don't have a GPU
+# )
+
+# # Conversion: 1 eV = 23.0605 kcal/mol
+# EV_TO_KCAL = 23.0605
+
+# print(f"{'Rxn ID':<8} | {'Barrier (eV)':<12} | {'Barrier (kcal/mol)':<18}")
+# print("="*50)
+
+# # 2. Loop through all reaction folders (assuming 1 to 20)
+# for i in range(1, 21):
+#     rxn_id = str(i)
+#     energies = {}
+
+#     # Check both Reactant and Transition State
+#     for state in ['react', 'ts']:
+#         # Construct the path: e.g., "1/react/POSCAR"
+#         path = os.path.join(rxn_id, state, 'POSCAR')
+
+#         if os.path.exists(path):
+#             try:
+#                 # Read the geometry
+#                 atoms = read(path)
+
+#                 # Attach the calculator we loaded earlier
+#                 atoms.calc = macemp
+
+#                 # Calculate Potential Energy
+#                 e_pot = atoms.get_potential_energy()
+#                 energies[state] = e_pot
+
+#                 # OPTIONAL: Write to file to save progress
+#                 # This overwrites ('w') to prevent duplicate lines if you re-run
+#                 output_file = os.path.join(rxn_id, state, 'energy-mace')
+#                 with open(output_file, "w") as f:
+#                     print(e_pot, file=f)
+
+#             except Exception as e:
+#                 print(f"Error calculating {path}: {e}")
+#         else:
+#             # Silently skip if folder doesn't exist 
+#             # (useful if you only have some folders)
+#             pass
+
+#     # 3. Calculate and Print Barrier
+#     if 'react' in energies and 'ts' in energies:
+#         barrier_ev = energies['ts'] - energies['react']
+#         barrier_kcal = barrier_ev * EV_TO_KCAL
+#         print(f"{rxn_id:<8} | {barrier_ev:.4f}       | {barrier_kcal:.4f}")
+#     else:
+#         print(f"{rxn_id:<8} | {'MISSING DATA':<30}")


can we remove this?

joehart2001 · 2026-05-14T13:30:28Z

Hey sorry for the long delay in getting on top of this! I've rebased the PR, changed the naming to CRBH20, did the pre-commit formatting, and also I removed some raw data from the repo. I'm also attaching the data zip file for the S3 bucket. Please review and let me know if there's anything else that needs to be done! CRBH20.zip

I've uplaoded and written a code suggestion for the download. If you accept hte code suggestion, remember to fetch and pull the changes to your local machine before doing any edits, otherwise they will be overwritten. Thanks!

joehart2001 · 2026-05-14T13:33:22Z

+
+# ml_peg imports
+from ml_peg.analysis.utils.decorators import build_table, plot_parity
+from ml_peg.analysis.utils.utils import build_d3_name_map, load_metrics_config, mae


Suggested change

from ml_peg.analysis.utils.utils import build_d3_name_map, load_metrics_config, mae

from ml_peg.analysis.utils.utils import build_dispersion_name_map, load_metrics_config, mae

joehart2001 · 2026-05-14T13:33:35Z

+from ml_peg.app import APP_ROOT
+from ml_peg.calcs import CALCS_ROOT
+from ml_peg.models.get_models import get_model_names
+from ml_peg.models.models import current_models


Suggested change

from ml_peg.models.models import current_models

from ml_peg.models import current_models

joehart2001 · 2026-05-14T13:33:45Z

+
+# --- Configuration ---
+MODELS = get_model_names(current_models)
+D3_MODEL_NAMES = build_d3_name_map(MODELS)


Suggested change

D3_MODEL_NAMES = build_d3_name_map(MODELS)

D3_MODEL_NAMES = build_dispersion_name_map(MODELS)

joehart2001 · 2026-05-14T13:34:08Z

+)
+from ml_peg.app.utils.load import read_plot
+from ml_peg.models.get_models import get_model_names
+from ml_peg.models.models import current_models


Suggested change

from ml_peg.models.models import current_models

from ml_peg.models import current_models

ElliottKasoar · 2026-05-14T14:43:31Z

+OUT_PATH = Path(__file__).parent / "outputs"
+
+# Unit conversion: 1 eV = 23.0605 kcal/mol
+EV_TO_KCAL = 23.0605


Suggested change

EV_TO_KCAL = 23.0605

EV_TO_KCAL = units.mol / units.kcal

Assuming you also add the units import above. In general it's better to import defined values than defining this ourselves everywhere

ElliottKasoar · 2026-05-14T14:43:44Z

+from pathlib import Path
+from typing import Any
+
+from ase.io import read, write


Suggested change

from ase.io import read, write

from ase import units

from ase.io import read, write

See below

ElliottKasoar · 2026-05-14T14:45:18Z

+
+    # 1. Initialize the Calculator
+    # The ml-peg wrapper handles device selection (cuda/cpu) and loading
+    calc = model.get_calculator()


Suggested change

calc = model.get_calculator()

calc = model.get_calculator(precision="high")

This is the default, but it's good to specify that we want 'high' (float64, usually) precision when we get the calculator

ElliottKasoar · 2026-05-14T14:49:43Z

+    print("=" * 50)
+
+    # 2. Loop through Reaction Folders (1 to 20)
+    for i in range(1, 21):


It would be nice to wrap this in use tqdm here to show the progress as we go through these iterations. This also potentially reduces the need for the print statements - we don't really need to print the barrier to the terminal, but is useful to see progress

ElliottKasoar · 2026-05-14T14:50:45Z

+                atoms.calc = calc
+
+                # Compute Energy
+                e_pot = atoms.get_potential_energy()


This isn't something we strictly enforce yet, but to get ahead of things, it would be useful to wrap this in a try...except to catch any errors e.g. if a model doesn't support an element in the dataset. If it fails, we'd probably want the energy to be set to NaN

ElliottKasoar · 2026-05-14T14:52:13Z

+                atoms.calc = calc
+
+                # Compute Energy
+                e_pot = atoms.get_potential_energy()


Suggested change

atoms.calc = calc

# Compute Energy

e_pot = atoms.get_potential_energy()

atoms.calc = calc

atoms.info.setdefault("charge", 0)

atoms.info.setdefault("spin", 1)

# Compute Energy

e_pot = atoms.get_potential_energy()

Unless these are read in from the input files, generally we need to set these, else certain models fail

ElliottKasoar · 2026-05-14T14:53:57Z

+# If the file doesn't exist, we provide defaults, but usually it should exist.
+try:
+    DEFAULT_THRESHOLDS, DEFAULT_TOOLTIPS, DEFAULT_WEIGHTS = load_metrics_config(
+        METRICS_CONFIG_PATH
+    )
+except FileNotFoundError:
+    # Fallback defaults if metrics.yml is missing
+    DEFAULT_THRESHOLDS = {"MAE": [1.0, 5.0]}  # Green < 1.0, Red > 5.0
+    DEFAULT_TOOLTIPS = {"MAE": "Mean Absolute Error"}
+    DEFAULT_WEIGHTS = {}


Suggested change

# If the file doesn't exist, we provide defaults, but usually it should exist.

try:

DEFAULT_THRESHOLDS, DEFAULT_TOOLTIPS, DEFAULT_WEIGHTS = load_metrics_config(

METRICS_CONFIG_PATH

)

except FileNotFoundError:

# Fallback defaults if metrics.yml is missing

DEFAULT_THRESHOLDS = {"MAE": [1.0, 5.0]} # Green < 1.0, Red > 5.0

DEFAULT_TOOLTIPS = {"MAE": "Mean Absolute Error"}

DEFAULT_WEIGHTS = {}

DEFAULT_THRESHOLDS, DEFAULT_TOOLTIPS, DEFAULT_WEIGHTS = load_metrics_config(

METRICS_CONFIG_PATH

)

If the metrics.yml doesn't exist, I think we want it to raise an error, so we can fix it. This makes it harder to catch

ElliottKasoar · 2026-05-14T14:54:59Z

+REF_BARRIERS_KCAL = {
+    1: 1.7194 * 23.0605,
+    2: 1.9241 * 23.0605,
+    3: 1.7499 * 23.0605,
+    4: 1.8238 * 23.0605,
+    5: 1.7237 * 23.0605,
+    6: 1.5653 * 23.0605,
+    7: 1.0911 * 23.0605,
+    8: 1.8983 * 23.0605,
+    9: 1.5477 * 23.0605,
+    10: 1.7115 * 23.0605,
+    11: 1.7379 * 23.0605,
+    12: 2.0361 * 23.0605,
+    13: 1.8739 * 23.0605,
+    14: 1.9760 * 23.0605,
+    15: 1.8865 * 23.0605,
+    16: 1.5741 * 23.0605,
+    17: 1.2587 * 23.0605,
+    18: 1.7497 * 23.0605,
+    19: 1.6989 * 23.0605,
+    20: 1.7654 * 23.0605,
+}


As in the calc, I would set a conversion factor using units.mol / units.kcal (probably as a single constant, not written out each time).

It's also ok to stick to eV if you want, it doesn't have to be kcal/mol

ElliottKasoar · 2026-05-14T14:56:36Z

+
+
+def numeric_sort_key(filepath: Path):
+    """Sort helper to ensure files 1, 2, ... 10 come in numerical order, not alpha."""


Alpha? As in alphanumerically? I think I'd either use the full word, or leave is sorting numerically, I think it's clear enough

ElliottKasoar · 2026-05-14T14:57:22Z

+    # --- DEBUGGING START ---
+    print(f"\nDEBUG: CALC_PATH is set to: {CALC_PATH.resolve()}")
+    print(f"DEBUG: OUT_PATH is set to:  {OUT_PATH.resolve()}")
+    # --- DEBUGGING END ---


I don't think we really need these debugging prints?

Suggested change

# --- DEBUGGING START ---

print(f"\nDEBUG: CALC_PATH is set to: {CALC_PATH.resolve()}")

print(f"DEBUG: OUT_PATH is set to: {OUT_PATH.resolve()}")

# --- DEBUGGING END ---

ElliottKasoar · 2026-05-14T14:59:28Z

+        if not model_dir.exists():
+            print(f"[MISSING] -> Skipped {model_dir}")
+            continue
+        print("[FOUND]")
+        # --- DEBUGGING END ---
+        if not model_dir.exists():
+            continue


Suggested change

if not model_dir.exists():

print(f"[MISSING] -> Skipped {model_dir}")

continue

print("[FOUND]")

# --- DEBUGGING END ---

if not model_dir.exists():

continue

if not model_dir.exists():

continue

I don't think we need debugging in this form. We will make it clearer which models were successful in other ways in future.

You also have duplicated if not model_dir.exists(): checks, so the second is never reached?

ElliottKasoar · 2026-05-14T15:01:11Z

+                # Handle missing data (e.g., if calc failed)
+                # For parity plots, lists must be equal length.
+                # We append None or NaN, though ml-peg might prefer dropping the point.
+                # Here we assume completeness or append 0.0 with a warning.
+                model_barriers.append(None)


Suggested change

# Handle missing data (e.g., if calc failed)

# For parity plots, lists must be equal length.

# We append None or NaN, though ml-peg might prefer dropping the point.

# Here we assume completeness or append 0.0 with a warning.

model_barriers.append(None)

model_barriers.append(np.nan)

Generally we're going to NaNs for missing/failed results

ElliottKasoar · 2026-05-14T15:02:22Z

+
+            # Extract ML Barrier (calculated in the previous script)
+            # stored as "barrier_kcal"
+            barrier_ml = reactant.info.get("barrier_kcal", 0.0)


Suggested change

barrier_ml = reactant.info.get("barrier_kcal", 0.0)

barrier_ml = reactant.info.get("barrier_kcal", np.nan)

It this key doesn't exist, we should be consistent with the above failure output, so either NaN or None if we didn't change above

ElliottKasoar · 2026-05-14T15:02:59Z

+
+            # Store reference energies (only once, during the first model loop)
+            if not ref_stored:
+                ref_val = REF_BARRIERS_KCAL.get(rxn_id, 0.0)


Suggested change

ref_val = REF_BARRIERS_KCAL.get(rxn_id, 0.0)

ref_val = REF_BARRIERS_KCAL.get(rxn_id, np.nan)

As above, although this shouldn't really happen. 0 will give confusing results

ElliottKasoar · 2026-05-14T15:06:16Z

+        results[model_name] = model_barriers
+
+        # Mark reference as stored so we don't duplicate it
+        if any(x is not None for x in model_barriers):


Generally it's easier to do this within if not ref_stored:, so we're not doing every checks for every model.

To be honest, if the reference is missing so gets set to None/NaN, I think something has probably gone wrong that continuing isn't necessarily useful anyway, so you could probably simplify REF_BARRIERS_KCAL.get(, as that's entirely in your control, and never set a None/NaN for the reference

ElliottKasoar · 2026-05-14T15:07:58Z

+            # Filter out None values in case of failed calculations
+            y_true = []
+            y_pred = []
+            for r, p in zip(reaction_barriers["ref"], reaction_barriers[model_name], strict=False):
+                if r is not None and p is not None:
+                    y_true.append(r)
+                    y_pred.append(p)


We don't want to filter out None/NaNs really. This is also an advantage to NaN over None as generally it's better handled - the mean of something with a NaN in it is NaN.

If all but one calculation failed, we don't want the score to be that error, the test has failed. In future, with element filtering, we'll use partial results more effectively

ElliottKasoar · 2026-05-14T15:09:07Z

+    good: 1.0   # Any error below 1.0 will be Green (Chemical Accuracy)
+    bad: 3.0    # Any error above 3.0 will be Red


Suggested change

good: 1.0 # Any error below 1.0 will be Green (Chemical Accuracy)

bad: 3.0 # Any error above 3.0 will be Red

good: 1.0

bad: 3.0

You don't really need to explain what these do, and the colours are also not necessarily correct.

Genuine question: How much have you considered these thresholds/do you understand what they should represent?

ElliottKasoar · 2026-05-14T15:09:26Z

@@ -0,0 +1,7 @@
+metrics:
+  MAE:
+    tooltip: "Mean Absolute Error (kcal/mol)"


Suggested change

tooltip: "Mean Absolute Error (kcal/mol)"

tooltip: "Mean Absolute Error"

This should be taken automatically fro the unit

ElliottKasoar · 2026-05-14T15:09:57Z

+# Placeholder docs link
+DOCS_URL = "https://github.com/ddmms/ml-peg"


Can you add documentation for this, and then update this please?

ElliottKasoar · 2026-05-14T15:10:45Z

+        # Dash Asset Paths: The 'assets' folder is mounted at APP_ROOT/data
+        # So we construct the URL path starting from there.
+        structs = [
+            f"assets/reaction_barriers/CRBH20/{valid_model}/{f.name}" for f in files


Suggested change

f"assets/reaction_barriers/CRBH20/{valid_model}/{f.name}" for f in files

f"/assets/reaction_barriers/CRBH20/{valid_model}/{f.name}" for f in files

Minor fix due to backend changes

ElliottKasoar · 2026-05-14T15:11:24Z

+    # Initialize Dash
+    # IMPORTANT: We set assets_folder to 'ml_peg/app/data' so it finds your JSONs/XYZs


Suggested change

# Initialize Dash

# IMPORTANT: We set assets_folder to 'ml_peg/app/data' so it finds your JSONs/XYZs

We don't really need these sorts of comments

ElliottKasoar · 2026-05-14T15:11:48Z

+# Placeholder docs link
+DOCS_URL = "https://github.com/ddmms/ml-peg"
+
+# Point to where we saved the JSONs (ml_peg/app/data/reaction_barriers/CRBH20)


Suggested change

# Point to where we saved the JSONs (ml_peg/app/data/reaction_barriers/CRBH20)

This doesn't really add anything

ElliottKasoar · 2026-05-14T15:13:01Z

I believe this has been uploaded, so we shouldn't need this as part of the pull request anymore

ElliottKasoar added the new benchmark Proposals and suggestions for new benchmarks label Feb 6, 2026

Quantumplations added 7 commits May 14, 2026 10:51

added test benchmark

c72f1fc

huh

320ea87

Added Flaviano's CRBH20 dataset files, in addition to a completed cal…

280efc3

…c_rxn_barriers.py script

Added Flaviano's CRBH20 dataset, also wrote the calc_rxn_barriers.py …

453c055

…script

added analysis_rxn_barriers.py script with unit test

3dda997

more commits

588639d

added the app_rxn_barriers.py script to setup Dash board, verified co…

cf6700d

…rrect display of model MAEs in browser

Quantumplations force-pushed the main branch from 5944a4e to cf6700d Compare May 14, 2026 09:57

Quantumplations added 4 commits May 14, 2026 11:07

Rename rxn_barriers to CRBH20 and revert collateral files

5b50d99

style: apply automatic pre-commit formatting

b7d7cbc

chore: remove raw data directory for S3 hosting

30877f1

style: fix ruff linting errors and format docstrings

b381c25

joehart2001 reviewed May 14, 2026

View reviewed changes

ElliottKasoar reviewed May 14, 2026

View reviewed changes

		if hasattr(model, "add_d3_calculator"):
		calc = model.add_d3_calculator(calc)

	from ml_peg.models.models import current_models
	from ml_peg.models import current_models

	# ml_peg imports for model management
	from ml_peg.calcs.utils.utils import download_s3_data

+    data_dir = (
+        download_s3_data(
+            key="inputs/molecular/CRBH20/CRBH20.zip",
+            filename="CRBH20.zip",
+        )
+        / "CRBH20"
+    )

	from ml_peg.analysis.utils.utils import build_d3_name_map, load_metrics_config, mae
	from ml_peg.analysis.utils.utils import build_dispersion_name_map, load_metrics_config, mae

	D3_MODEL_NAMES = build_d3_name_map(MODELS)
	D3_MODEL_NAMES = build_dispersion_name_map(MODELS)

	from ase.io import read, write
	from ase import units
	from ase.io import read, write

	calc = model.get_calculator()
	calc = model.get_calculator(precision="high")



		def numeric_sort_key(filepath: Path):
		"""Sort helper to ensure files 1, 2, ... 10 come in numerical order, not alpha."""

	barrier_ml = reactant.info.get("barrier_kcal", 0.0)
	barrier_ml = reactant.info.get("barrier_kcal", np.nan)

	ref_val = REF_BARRIERS_KCAL.get(rxn_id, 0.0)
	ref_val = REF_BARRIERS_KCAL.get(rxn_id, np.nan)

		good: 1.0 # Any error below 1.0 will be Green (Chemical Accuracy)
		bad: 3.0 # Any error above 3.0 will be Red

	tooltip: "Mean Absolute Error (kcal/mol)"
	tooltip: "Mean Absolute Error"

		# Placeholder docs link
		DOCS_URL = "https://github.com/ddmms/ml-peg"

	f"assets/reaction_barriers/CRBH20/{valid_model}/{f.name}" for f in files
	f"/assets/reaction_barriers/CRBH20/{valid_model}/{f.name}" for f in files

		# Initialize Dash
		# IMPORTANT: We set assets_folder to 'ml_peg/app/data' so it finds your JSONs/XYZs

Conversation

Quantumplations commented Feb 4, 2026

Pre-review checklist for PR author

Summary

Linked issue

Progress

Testing

New decorators/callbacks

Uh oh!

Quantumplations commented May 14, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

joehart2001 May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

joehart2001 commented May 14, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ElliottKasoar May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ElliottKasoar May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ElliottKasoar May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ElliottKasoar May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

joehart2001 May 14, 2026 •

edited

Loading

ElliottKasoar May 14, 2026 •

edited

Loading

ElliottKasoar May 14, 2026 •

edited

Loading

ElliottKasoar May 14, 2026 •

edited

Loading

ElliottKasoar May 14, 2026 •

edited

Loading