#!/bin/bash
#SBATCH --job-name=tier2_bio
#SBATCH --partition=nova
#SBATCH --gres=gpu:1
#SBATCH --cpus-per-task=8
#SBATCH --mem=32G
#SBATCH --time=02:00:00
#SBATCH --output=bert_v6_contrastive/scripts/tier2_bio_%j.out
#SBATCH --error=bert_v6_contrastive/scripts/tier2_bio_%j.err

echo "=== Tier 2: Core Biology Embedding Analysis ==="
echo "Started: $(date)"
echo "Node: $(hostname)"

cd /work/ratul1/supantha/glycan-SD-VS/bert_training_v3/v3.1_cluster_training
source /work/ratul1/supantha/miniconda3/etc/profile.d/conda.sh
conda activate glycan_bert 2>/dev/null || source activate glycan_bert 2>/dev/null || true
pip install -q scipy 2>/dev/null

echo ""
echo "=== Step 1: Analyze V6 (Tier 2) ==="
python3 bert_v6_contrastive/scripts/analyze_core_biology.py --model v6

echo ""
echo "=== Step 2: Analyze V5 (Tier 2) ==="
python3 bert_v6_contrastive/scripts/analyze_core_biology.py --model v5

echo ""
echo "=== Step 3: V5 vs V6 Tier 2 Comparison ==="
python3 -c "
import json
v5 = json.load(open('bert_v6_contrastive/analysis/tier2_metrics_v5.json'))
v6 = json.load(open('bert_v6_contrastive/analysis/tier2_metrics_v6.json'))
print('=== Tier 2 Comparison Summary ===')
for key in sorted(set(list(v5.keys()) + list(v6.keys()))):
    if key in ('model', 'tier', 'structural_type_counts', 'branching_distribution', 'linkage_distribution'):
        continue
    v5v = v5.get(key, 'N/A')
    v6v = v6.get(key, 'N/A')
    if isinstance(v5v, (int, float)) and isinstance(v6v, (int, float)):
        delta = v6v - v5v
        print(f'  {key}: V5={v5v}, V6={v6v}, D={delta:+.4f}')
    else:
        print(f'  {key}: V5={v5v}, V6={v6v}')
"

echo ""
echo "=== Output Files ==="
ls -la bert_v6_contrastive/analysis/tier2_* bert_v6_contrastive/analysis/*structural* bert_v6_contrastive/analysis/*composition* bert_v6_contrastive/analysis/*branching* bert_v6_contrastive/analysis/*linkage* bert_v6_contrastive/analysis/*decoration* bert_v6_contrastive/analysis/*plantae* 2>/dev/null

echo ""
echo "Done! $(date)"