Synthetic sensor data generation with automatic imputation and model selection. A TSGM wrapper which automatically handles data ingestion, imputation and modelling part. Configuration is though a Claude API and also manual. Handles missing data and data type modelling automatically.
pip install synthflow
from synthflow import SynFlow
sf = SynFlow(mode="auto", api_key="sk-ant-...") sf.chat("accelerometer from a wind turbine, 500Hz, dropout during storms") sf.chat("generate")
from synthflow import SynFlow from synthflow.parser import SynConfig
config = SynConfig( domain="industrial", sensor_type="accelerometer", sampling_rate_hz=500, backend="tsgm", model="TimeVAE", ) sf = SynFlow(mode="manual", config=config, data="sensor_data.csv") result = sf.generate() print(result.data.head())