interpy-synth 0.1.1

Synthetic 5D→1D dataset generator used by interpy_bg and fivedreg_tf.

interpy_synth

Lightweight synthetic data generator for 5D → 1D regression workloads. Provides helper functions to create NumPy arrays or persist pickles with metadata. Used by both the NumPy (interpy_bg) and TensorFlow (fivedreg_tf) packages, but can be installed standalone.

Installation

pip install interpy-synth
# or from this repo:
# pip install git+https://github.com/barongracias/InterPyApp.git#egg=interpy-synth&subdirectory=backend/interpy_synth

Usage

from interpy_synth import synthetic_5d, synthetic_5d_pickle

X, y = synthetic_5d(1000, seed=42)
path = synthetic_5d_pickle("outputs_numpy/synth.pkl", n=1000, seed=42)

synthetic_5d returns float32 arrays shaped (n, 5) and (n, 1).
synthetic_5d_pickle writes a pickle with X, y, and metadata (feature names, seed, timestamp).