Add separate section for Parquet tools

awesome-spark · Aug 3, 2016 · b304bf9 · b304bf9
1 parent b803e6b
commit b304bf9
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/06_data_preparation.md b/06_data_preparation.md
@@ -190,7 +190,7 @@ model.transform(df).schema[-1].metadata
 ##   'num_attrs': 2}}
 ```
 
-As for now PySpark doesn't support attaching metadata to a single column. It is possible though to use method similar to this one:
+As for now PySpark doesn't support attaching metadata to a single column. It is possible though, to use method similar to this one:
 
 ```python
 import json
@@ -257,6 +257,8 @@ spark.read.parquet("/tmp/foo").schema.headOption.map(_.metadata)
 // Option[org.apache.spark.sql.types.Metadata] = Some({"foo":"bar"})
 ```
 
+#### Accessing Metadata Directly
+
 Metadata can be also accessed directly using Parquet tools:
 
 ```scala