Getting Started

The fastest path from zero to a working Skyulf pipeline.

1. Install

pip install skyulf-core

Or install with all optional extras:

pip install skyulf-core[viz,eda,tuning,modeling-xgboost,preprocessing-imbalanced]

For editable installs and Docker, see Installation.

2. Minimal example

import pandas as pd
from skyulf.pipeline import SkyulfPipeline

df = pd.DataFrame({
    "age": [10, 20, None, 40, 50, 60, None, 80],
    "city": ["A", "B", "A", "C", "B", "A", "C", "B"],
    "target": [0, 1, 0, 1, 1, 0, 1, 0],
})

config = {
    "preprocessing": [
        {"name": "split", "transformer": "TrainTestSplitter",
         "params": {"test_size": 0.25, "random_state": 42,
                    "stratify": True, "target_column": "target"}},
        {"name": "impute", "transformer": "SimpleImputer",
         "params": {"columns": ["age"], "strategy": "mean"}},
        {"name": "encode", "transformer": "OneHotEncoder",
         "params": {"columns": ["city"], "drop_original": True}},
    ],
    "modeling": {
        "type": "random_forest_classifier",
        "params": {"n_estimators": 50, "random_state": 42},
    },
}

pipeline = SkyulfPipeline(config)
metrics = pipeline.fit(df, target_column="target")
print(metrics)

preds = pipeline.predict(df.drop(columns=["target"]))
print(preds.head())

3. What just happened?

1. TrainTestSplitter separated data into train/test sets (no leakage).
2. SimpleImputer learned the mean of age from training data only.
3. OneHotEncoder created dummy columns for city.
4. RandomForestClassifier trained on the processed training split.

4. Next steps

Goal	Page
Full train / evaluate / save / load workflow	Pipeline Quickstart
All 20 supported models and config keys	Configuration
Hyperparameter tuning (grid, random, Optuna)	Hyperparameter Tuning
Add your own custom nodes	Extending Skyulf-Core
Full platform setup (backend + UI)	Architecture