-
If the hive tables for storing vertices are not created, create them first. Substitute the table names in the bash scripts if necessary.
-
Execute the start.sh script. Change the IO path in the Spark scripts if necessary.
cd vertices/ cd preparation/ & sh modify_schema_for_vertex_raw.sh & sh modify_schema_for_vertex.sh & cd .. sh start.sh cd ../
-
If the hive table for storing edges is not created, create it first. Substitute the table name in the bash script if necessary.
-
Execute the start.sh script. Change the IO path in the Spark scripts if necessary.
cd edges/ cd preparation/ & sh modify_schema_for_edge.sh & cd .. sh start.sh cd ../
- If the hive tables for storing features are not created, create them first. Substitute the table names in the bash scripts accordingly.
- Execute the start.sh script. Change the IO path in the Spark scripts if necessary.
cd features/
cd preparation/ & sh modify_schema_for_ins_features.sh & sh modify_schema_for_dev_features.sh & cd ..
sh start.sh
cd ../
- Please refer to the README file in the algorithms sub directory