When testing (new) queries, one can use the browser or Eclipse.
Some thoughts on pros and cons (with examples at the bottom if applicable):
Browser
++ 1. Browser visualizes results automatically
++ 2. Browser has autocomplete for node/edge labels + properties
++ 3. Browser shows “query execution plan”
– 4. Browser can be quite slow in visualizing results of complex querie(s).
– 5. Browser timer is not that accurate
Eclipse
++ 6. Eclipse is fast in executing a query (without visualisation)
++ 7. Eclipses can time execution of queries very accurate
– 8. Obtaining results of Nodes with different properties needs quite some coding
Examples:
- To change the visualization in the GUI, see the extensive (Neo4j documentation)[https://neo4j.com/developer/neo4j-browser/] and example below.
- 1 Click on one of the Node labels to select it (top left corner of query results visualization window).
- 2 Select the visualization option you desire in the panel (bottom left corner of query results visualization window)
- A more complex query is the shortest path, especially when performed on a (large) set of Nodes. When executing shortest path, Neo4j can use two algorithms, a fast algorithm (bidirectional breadth-first search), or a much slower exhaustive depth-first search algorithm. Therefore it is important to plan how the shortest path is executed; calculating all paths for all nodes is computationally much more time consuming than only using a handful of nodes. Add the “EXPLAIN” term in front of a query, click run and in the result section on “plan” (see also image below).
