fastparquet_pandas_dataframe_parquet_read_write_quickstart.py

1import numpy as np
2import pandas as pd
3from fastparquet import write, ParquetFile
4
5# Create a sample DataFrame
6df = pd.DataFrame({'a': np.arange(10),
7                   'b': np.random.randn(10)})
8
9# Write the DataFrame to a Parquet file
10write('outfile.parquet', df)
11
12# Read the Parquet file back into a DataFrame
13pf = ParquetFile('outfile.parquet')
14df2 = pf.to_pandas()
15
16# Alternatively, use the pandas integration
17df3 = pd.read_parquet('outfile.parquet', engine='fastparquet')