Scoring System¶

ASCICat uses a three-pillar scoring system to evaluate catalysts comprehensively.

Overview¶

Each catalyst receives three normalized scores:

Score	Symbol	Based On	Range
Activity	\(S_a\)	Adsorption energy \(\Delta E\)	[0, 1]
Stability	\(S_s\)	Surface energy \(\gamma\)	[0, 1]
Cost	\(S_c\)	Material cost \(C\)	[0, 1]

These scores are combined into the unified ASCI metric:

\[\phi_{ASCI} = w_a \cdot S_a + w_s \cdot S_s + w_c \cdot S_c\]

Scoring Philosophy¶

Why Normalize to [0, 1]?¶

Comparability: All scores use the same scale
Interpretability: 0 = worst, 1 = best (intuitive)
Weighted combination: Weights directly reflect importance
Bounded output: ASCI is also in [0, 1]

Data-Driven vs. Fixed Ranges¶

Approach	Stability/Cost	Pros	Cons
Data-driven	min/max from dataset	Full [0,1] range used	Scores not comparable across datasets
Fixed ranges	Predetermined values	Cross-study comparison	Some scores may cluster

ASCICat uses data-driven normalization by default for stability and cost scores.

Detailed Documentation¶

Activity Scoring

Sabatier principle implementation with linear and Gaussian methods

Activity Score
Stability Scoring

Surface energy-based stability assessment

Stability Score
Cost Scoring

Logarithmic economic viability scoring

Cost Score
ASCI Integration

Combining scores into the unified metric

ASCI Integration

Quick Reference¶

Activity Score¶

from ascicat.scoring import score_activity

# Linear method (default)
S_a = score_activity(
    delta_E=-0.25,       # Your adsorption energy
    optimal_E=-0.27,     # Sabatier optimum
    width=0.15,          # Tolerance
    method='linear'
)
# Returns: 0.867

Stability Score¶

from ascicat.scoring import score_stability

# Data-driven normalization
gammas = [0.52, 1.0, 2.0, 3.0, 4.5]
S_s = score_stability(gammas)
# Returns: [1.0, 0.88, 0.63, 0.38, 0.0]

Cost Score¶

from ascicat.scoring import score_cost

# Logarithmic normalization
costs = [10, 100, 1000, 10000]
S_c = score_cost(costs)
# Returns: [1.0, 0.67, 0.33, 0.0]

Combined ASCI¶

from ascicat.scoring import calculate_asci

phi = calculate_asci(
    activity_score=0.85,
    stability_score=0.70,
    cost_score=0.90,
    w_a=0.33, w_s=0.33, w_c=0.34
)
# Returns: 0.818

Validation¶

All scores are validated:

Range checking: Inputs must be physical (e.g., cost > 0)
Weight validation: Must sum to 1.0
Output clipping: Scores guaranteed in [0, 1]

from ascicat.config import validate_weights

# This passes
validate_weights(0.33, 0.33, 0.34)

# This raises ValueError
validate_weights(0.5, 0.3, 0.3)  # Sum = 1.1