Modified based on feedback - removing pandas

removing pandas requirement and using Pyspark FillNa directly

examples/transforms/CustomTransform_FillEmptyStringsInAColumn.py

@@ -3,7 +3,6 @@
 from pyspark.sql.dataframe import DataFrame
 from pyspark.sql.types import StructType
 from pyspark.context import SparkContext
-import pandas as pd
 from pyspark.sql.functions import col,isnan, when, count, regexp_replace
 
 def fill_empty_null_values_txn(
@@ -16,12 +15,9 @@ def fill_empty_null_values_txn(
             _dyf = DynamicFrame.fromDF(modifiedDF, self.glue_ctx, self.name)
             return _dyf
         elif _df.filter(col(columnName).isNull()).count() > 0:  
-            _pdf = _df.toPandas()
-            _pdf[columnName] = _pdf[columnName].fillna(newValue)
-            modifiedDF = gluectx.spark_session.createDataFrame(_pdf) 
-            _dyf = DynamicFrame.fromDF(modifiedDF, self.glue_ctx, self.name)
+            _df = _df.fillna(value=newValue, subset=[columnName])
+            _dyf = DynamicFrame.fromDF(_df, self.glue_ctx, self.name)
             return _dyf
         return self
 
-
-DynamicFrame.fill_empty_null_values_txn = fill_empty_null_values_txn
+DynamicFrame.fill_empty_null_values_txn = fill_empty_null_values_txn