Fuzzy Kulli-Basava Index (FKB₁*) Computation Package for Isomer Discrimination and QSPR Analysis
Authors/Creators
Description
**Python implementation of the generalized fuzzy Kulli-Basava index FKB₁* for fuzzy graphs**
This package implements the complete computational framework for the fuzzy Kulli-Basava index (FKB₁*) as described in "A Fuzzy Graph-Based Kulli–Basava Descriptor for Isomer Discrimination and Physicochemical Property Prediction".
## Features
- Computation of FKB₁* for arbitrary fuzzy graphs (Algorithm 1)
- Computation of comparative fuzzy indices: fuzzy first Zagreb index (FZ₁) and fuzzy Randić index (FR)
- Non-uniform membership assignment by carbon substitution class
- Closed-form formulas for five graph families:
- Paths Pₙ (Theorem 3.16)
- Cycles Cₙ (Theorem 3.17)
- Complete graphs Kₙ (Theorem 3.18)
- Complete bipartite graphs K_{a,b} (Theorem 3.19)
- Wheels Wₙ (Theorem 3.20)
- Complete molecular graph generators for:
- Hexane isomers (C₆H₁₄, 5 isomers)
- Heptane isomers (C₇H₁₆, 9 isomers)
- Octane isomers (C₈H₁₈, 18 isomers)
- QSPR regression analysis for thermodynamic properties:
- Acentric factor
- Entropy (S)
- Enthalpy of vaporization (ΔH_vap)
- Enthalpy of vaporisation (H_vap)
- Isomer discrimination analysis with degeneracy detection
- Two-descriptor enthalpy models with first Zagreb index (M₁)
## Reproduces Results
The code reproduces all results reported in the manuscript:
- Perfect isomer discrimination (d̄ = 1.000) for C₆ and C₇
- R² = 0.9681 for octane acentric factor prediction
- Q²_LOO = 0.9587 for octane acentric factor
- Resolution of three octane degenerate pairs using non-uniform memberships
- Two-descriptor enthalpy models achieving R² > 0.94
## Dependencies
- Python 3.7+
- NumPy 1.19.0+
## Usage Example
```python
from fkb1_star import *
# Define graph edges and uniform membership
edges = [(0,1), (1,2), (2,3)]
memberships = compute_uniform_membership(edges, 0.8)
# Compute FKB₁*
fkb = compute_fkb1_star(edges, memberships, 4)
print(f"FKB₁* = {fkb:.4f}")
Files
Files
(25.7 kB)
| Name | Size | |
|---|---|---|
|
md5:fa55ac23240814a0297f8719838a7538
|
25.7 kB | Download |