Create a time series from a CSV or parquet file without importing the content

Seddryck / Tseesecake

Tseesecake is a lightweight proxy for time-series query engines, supporting multiple database engines and storage providers, with a SQL dialect dedicated to time series

https://seddryck.github.io/Tseesecake

Apache License 2.0

2 stars 1 forks source link

Create a time series from a CSV or parquet file without importing the content #28

Open Seddryck opened 1 year ago

Seddryck commented 1 year ago

We should create a view that is providing this kind of query

CREATE OR REPLACE VIEW WindEnergy (Instant, WindPark, Producer, Forecasted, Produced) AS ...

For CSV:

SELECT * FROM read_csv('flights.csv', delim='|', header=True, 
columns={'FlightDate': 'DATE', 'UniqueCarrier': 'VARCHAR', 'OriginCityName': 'VARCHAR', 'DestCityName': 'VARCHAR'});

For Parquet:

SELECT * FROM read_parquet(['folder1/*.parquet','folder2/*.parquet']);

Seddryck commented 1 year ago