I just wrote a blog post on how to convert a CSV to Parquet with Go and thought about how wonderful it would be to play with Parquet files in a Go DataFrame ;)
I doesn't look like the readCSV code is that complex.
Do you think parquet-go can be used to add this feature relatively easily?