I have an S3 bucket which contains parquet files.

I need to analyse that parquet file and create the required table in Redshift serverless.

import pyarrow.parquet as pq
df = pq.read_table(f"s3://{bucket_name}/{s3_path}").to_pandas()
table_create_statement = pd.io.sql.get_schema(df, table_name)

Using the above code, I was able to get the create statement for the dataframe as table.
But the df columns contains numbers that are of 38 digits. where as the create table statement that I got contain the INTEGER data type which has numeric precision of 32 for that column instead of NUMERIC(38, 0) or DECIMAL(38, 0).
And for some columns, the table create statement have boolean type, where as the actual data contains numbers or strings – TRUE or FALSE

How do I get the table create statement that’s compatible for Redshift serverless?