[BUG]:When returning a StringDataFrameColumn from the apply method, a java.lang.IllegalArgumentException is thrown

[wzl-bxg]

Describe the bug
When returning a StringDataFrameColumn from the apply method, a java.lang.IllegalArgumentException is thrown.

To Reproduce

Steps to reproduce the behavior:
1.Implement an apply method that returns a StringDataFrameColumn.
2. Execute the code that triggers the apply method.
3. Observe that a java.lang.IllegalArgumentException is thrown at runtime

Minimal reproducible example:

• Spark version: 3.3.4
• Microsoft.Spark version: 2.3.0
• Microsoft.Spark.Worker: Microsoft.Spark.Worker.net8.0.win-x64-2.3.0

using Microsoft.Data.Analysis;
using Microsoft.Spark.Sql;
using Microsoft.Spark.Sql.Types;
using FxDataFrame = Microsoft.Data.Analysis.DataFrame;

namespace MySparkApp
{
class GroupApply
{
static void Main(string[] args)
{
SparkSession spark =
SparkSession
.Builder()
.AppName("group_apply_sample")
.GetOrCreate();

        var df = spark.CreateDataFrame(new[]
        {
            new GenericRow(new object[] { 10, 2, 5.0, "this is description one" }),
            new GenericRow(new object[] { 150, 10, 1.0, "this is description two" }),
            new GenericRow(new object[] { 150, 3, 1.0 , "this is description three"}),
        },
        new StructType(new[]
        {
            new StructField("ProductId", new IntegerType()),
            new StructField("Qty", new IntegerType()),
            new StructField("Price", new DoubleType()),
            new StructField("Description", new StringType())
        }));

        var returnSchema = new StructType(new[]
        {
            new StructField("Description", new StringType())
        });

        var result = df
            .GroupBy("ProductId")
            .Apply(returnSchema, dataFrame =>
            {
                var descriptions = new List<string>();

                foreach (var row in dataFrame.Rows)
                {
                    int productId = (int)row["ProductId"];
                    string description = (string)row["Description"];

                    if (productId > 10)
                    {
                        descriptions.Add(description + productId);
                    }
                }

                var descriptionColumn = new StringDataFrameColumn("Description", descriptions);

                var resultFrame = new FxDataFrame(descriptionColumn);

                return resultFrame;
            });

        result.Show();

        spark.Stop();
    }
}

}

Expected behavior
The apply method should successfully return a StringDataFrameColumn without throwing an exception.

Screenshots

Desktop (please complete the following information):

• OS: [Windows 10]
• Spark Version [3.3.4]
• Microsoft.Spark Version [2.3.0]
• Microsoft.Spark.Worker [Microsoft.Spark.Worker.net8.0.win-x64-2.3.0]

Additional context
This issue only occurs when the return type of the apply method is StringDataFrameColumn. Returning Int32DataFrameColumn and DoubleDataFrameColumn types will work as expected