file

Type Members

1. case class FileCsvGraphDataSource(rootPath: String)(implicit session: CAPSSession) extends CAPSPropertyGraphDataSource with Product with Serializable

   

   Loads a graph stored in CSV format from the local file system.

   Loads a graph stored in CSV format from the local file system. The CSV files must be stored following this schema:

   # Nodes

   • all files describing nodes are stored in a sub folder called "nodes"
   • create one file for each possible label combination that exists in the data. This means that a node can only be present in one file. Example: All nodes with labels :Person:Employee are in a single file and all nodes that have label :Person are stored in another file. A node that appears in :Person:Employee CANNOT appear again in the file for :Person.
   • for every node csv file create a schema file called FILE_NAME.csv.SCHEMA
   • for information about the structure of the node schema file see org.opencypher.spark.impl.io.hdfs.CsvNodeSchema

   # Relationships

   • all files describing relationships are stored in a sub folder called "relationships"
   • create one csv file per relationship type
   • for every relationship csv file create a schema file called FILE_NAME.csv.SCHEMA
   • for information about the structure of the relationship schema file see org.opencypher.spark.impl.io.hdfs.CsvRelSchema

   rootPath

   path to the folder containing the nodes/relationships folders

   session

   CAPS Session