TigerGraph Ecosystem
Key Folders:
-
etl - The Connector Ecosystem
- dist_split - utility to split large data files for loading onto a distributed system
- kafka-connect - kafka connector
- spark_loading - This program uses Spark SQL to load parquet from Amazon S3 and store as CSV locally or in S3.
- tg_connect - Contains two utilities to read csv files from Hadoop or records from a RDBMS and then posts to TigerGraph.
- tg-java-driver - JDBC Type 4 Driver
-
graph_algorithms - moved to https://github.com/tigergraph/gsql-graph-algorithms
-
gurus_scripts - Data and GSQL scripts so users can replicate the demostrations in our Graph Gurus webinar series
- comm_dect_demo
- fraud_detection_demo
- geospatial_search
- loop_detection_demo
- movie_recommendation
- network_IT_resource
- pagerank_demo
- pattern_match
- temporal_data