In [9]:

#r "nuget:Parquet.Net"

In [10]:

#r "nuget:System.IO"

In [15]:

using System.Collections.Generic;
using System.IO;
using System.Linq;
using Parquet; 
using Parquet.Data;

In [16]:

//create data columns with schema metadata and the data you need
var idColumn = new DataColumn(
   new DataField<int>("id"),
   new int[] { 1, 2 });

var cityColumn = new DataColumn(
   new DataField<string>("city"),
   new string[] { "London", "Derby" });

// create file schema
var schema = new Schema(idColumn.Field, cityColumn.Field);

In [21]:

Stream ms = new System.IO.MemoryStream();

using (var parquetWriter = new ParquetWriter(schema, ms))
{
  // create a new row group in the file
  using (ParquetRowGroupWriter groupWriter = parquetWriter.CreateRowGroup())
  {
     groupWriter.WriteColumn(idColumn);
     groupWriter.WriteColumn(cityColumn);
  }
}

In [22]:

ms.Position = 0;

In [25]:

// open parquet file reader
using (var parquetReader = new ParquetReader(ms))
{
  // get file schema (available straight after opening parquet reader)
  // however, get only data fields as only they contain data values
  DataField[] dataFields = parquetReader.Schema.GetDataFields();

  // enumerate through row groups in this file
  for(int i = 0; i < parquetReader.RowGroupCount; i++)
  {
     // create row group reader
     using (ParquetRowGroupReader groupReader = parquetReader.OpenRowGroupReader(i))
     {
        // read all columns inside each row group (you have an option to read only
        // required columns if you need to.
        DataColumn[] columns = dataFields.Select(groupReader.ReadColumn).ToArray();

        // get first column, for instance
        DataColumn firstColumn = columns[0];

        // .Data member contains a typed array of column data you can cast to the type of the column
        Array data = firstColumn.Data;
        int[] ids = (int[])data;
        foreach (int id in ids)
        { 
            Console.WriteLine(id);
        }
     }
  }
}

1
2

In [ ]: