evo-mars-grid-search
SkillSearchParallel grid search over DBSCAN hyperparameters for Mars cloud clustering. Aggregates per-image F1 and delta scores with proper averaging semantics.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the evo-mars-grid-search skill
What this skill tells your AI
The instructions your AI receives, as published by openlair/openskill in tasks-evolved/mars-clouds-clustering/environment/skills/evo-mars-grid-search/SKILL.md and read by ahel’s review.
Parallel grid search over DBSCAN hyperparameters.
Key Functions
load_data(citsci_path, expert_path)- Load CSVs, return grouped dictsevaluate_hyperparameter_combo(ms, eps, sw, images, citsci_dict, expert_dict)- Eval one comborun_grid_search(unique_images, citsci_dict, expert_dict, n_jobs)- Full parallel grid searchfilter_results(results_df, min_f1=0.5)- Filter to F1 > 0.5
Usage
import sys
sys.path.insert(0, '/app/environment/skills/evo-mars-grid-search/scripts')
from utils import load_data, run_grid_search, filter_results
_, _, images, citsci_dict, expert_dict = load_data('citsci.csv', 'expert.csv')
results = run_grid_search(images, citsci_dict, expert_dict)
filtered = filter_results(results)
Averaging Rules
- Loop over ALL unique images from expert dataset
- Images with no citsci points: F1=0.0, delta=NaN
- F1 average includes all zeros
- Delta average excludes NaN values
- Filter: only keep F1 > 0.5
Signals
- GitHub stars
- 89
- Forks
- 4
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
evo-mars-grid-search- Source
- github.com/openlair/openskill