Skip to content

Commit 9565fa6

Browse files
JohTJohT
committed
Visualize longest paths with GraphViz
1 parent 2e28ac6 commit 9565fa6

File tree

4 files changed

+187
-1
lines changed

4 files changed

+187
-1
lines changed

‎cypher/Path_Finding/Path_Finding_6_Longest_paths_contributors_for_graphviz.cypher‎

Lines changed: 73 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,73 @@
1+
// Path Finding - Longest path - Stream - List all dependencies for nodes contributing to longest paths and highlight those paths in the Visualization with GraphViz.
2+
3+
// Gather global statistics about dependency weights and levels for normalization and node details
4+
MATCH (sourceNodeForStatistics)-[dependencyForStatistics:DEPENDS_ON]->(targetNodeForStatistics)
5+
WHERE $dependencies_projection_node IN LABELS(sourceNodeForStatistics)
6+
AND $dependencies_projection_node IN LABELS(targetNodeForStatistics)
7+
WITH min(dependencyForStatistics[$dependencies_projection_weight_property]) AS minWeight
8+
,max(dependencyForStatistics[$dependencies_projection_weight_property]) AS maxWeight
9+
,max(targetNodeForStatistics.maxDistanceFromSource) AS maxLevel
10+
WITH *, 1.0 / toFloat(maxWeight - minWeight) AS weightNormalizationFactor
11+
WITH { minWeight: minWeight, maxLevel: maxLevel, weightNormalizationFactor: weightNormalizationFactor } AS statistics
12+
// -> Main call to execute "longest path" algorithm
13+
CALL gds.dag.longestPath.stream($dependencies_projection + '-cleaned')
14+
YIELD index, totalCost, path
15+
WITH *
16+
// Sort longest paths by their length descending and - if equal - by their index ascending
17+
ORDER BY totalCost DESC, index ASC
18+
// Collect all results of the longest path search as well as all nodes of the longest paths
19+
WITH statistics
20+
,collect({index: index, distance: toInteger(totalCost), path: path}) AS longestPaths
21+
,collect(nodes(path)) AS allLongestPathNodes
22+
// Flatten and deduplicate the list of all nodes that contribute to at least one longest path
23+
UNWIND allLongestPathNodes AS longestPathNodes
24+
UNWIND longestPathNodes AS longestPathNode
25+
WITH statistics
26+
,longestPaths
27+
,collect(DISTINCT longestPathNode) AS allDistinctLongestPathNodes
28+
// Iterate over all longest paths
29+
UNWIND longestPaths AS longestPath
30+
WITH statistics
31+
,longestPaths, allDistinctLongestPathNodes
32+
,[ singleRelationship IN relationships(longestPath.path) | [startNode(singleRelationship), endNode(singleRelationship)] ] AS allLongestPathStartAndEndNodeTuples
33+
,[ singleRelationship IN relationships(longestPaths[0].path) | [startNode(singleRelationship), endNode(singleRelationship)] ] AS longestPathStartAndEndNodeTuples
34+
,longestPath.index AS index
35+
,longestPath.distance AS distance
36+
// -> Main query of all dependencies of nodes contributing to the longest paths
37+
MATCH (source)-[dependency:DEPENDS_ON]->(target)
38+
WHERE $dependencies_projection_node IN labels(source)
39+
AND $dependencies_projection_node IN labels(target)
40+
// Dependent nodes need to be part of at least one longest paths
41+
AND (source IN allDistinctLongestPathNodes AND target IN allDistinctLongestPathNodes)
42+
WITH statistics.maxLevel AS maxLevel
43+
,statistics.minWeight AS minWeight
44+
,statistics.weightNormalizationFactor AS weightNormalizationFactor
45+
,count(index) AS numberOfLongestPathsPassing
46+
,max(distance) AS lengthOfLongestPathPassing
47+
,dependency
48+
,source
49+
,target
50+
// If there is at least one longest path passing through the dependency then "contributesToALongestPath" is true
51+
,([source, target] IN allLongestPathStartAndEndNodeTuples) AS contributesToALongestPath
52+
,([source, target] IN longestPathStartAndEndNodeTuples) AS isPartOfLongestPath
53+
WITH *, dependency[$dependencies_projection_weight_property] AS weight
54+
WITH *, toFloat(weight - minWeight) * weightNormalizationFactor AS normalizedWeight
55+
WITH *, round((normalizedWeight * 5) + 1, 2) AS penWidth
56+
WITH *, source.name + "\\n(level " + source.maxDistanceFromSource + "/" + maxLevel + ")" AS startNodeTitle
57+
WITH *, target.name + "\\n(level " + target.maxDistanceFromSource + "/" + maxLevel + ")" AS endNodeTitle
58+
// The longest path will be highlighted in red.
59+
WITH *, CASE WHEN isPartOfLongestPath THEN "; color=\"red\""
60+
// Dependencies contributing to the longest path will be highlighted in dark orange.
61+
WHEN contributesToALongestPath THEN "; color=\"darkorange\""
62+
ELSE "" END AS edgeColor
63+
WITH *, "[label=" + weight + "; penwidth=" + penWidth + edgeColor + "; ];" AS graphVizEdgeAttributes
64+
WITH *, "\"" + startNodeTitle + "\" -> \"" + endNodeTitle + "\" " + graphVizEdgeAttributes AS graphVizDotNotationLine
65+
RETURN DISTINCT graphVizDotNotationLine
66+
// Debugging
67+
// ,source.name
68+
// ,target.name
69+
// ,numberOfLongestPathsPassing
70+
// ,lengthOfLongestPathPassing
71+
// ,contributesToALongestPath
72+
// ,isPartOfLongestPath
73+
LIMIT 100

‎cypher/Path_Finding/Path_Finding_6_Longest_paths_for_graphviz.cypher‎

Lines changed: 35 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,35 @@
1+
// Path Finding - Longest path - Stream - Find the top 100 dependencies contributing to the longest paths for Visualization with GraphViz
2+
3+
MATCH (sourceNodeForStatistics)-[dependencyForStatistics:DEPENDS_ON]->(targetNodeForStatistics)
4+
WHERE $dependencies_projection_node IN LABELS(sourceNodeForStatistics)
5+
AND $dependencies_projection_node IN LABELS(targetNodeForStatistics)
6+
WITH min(dependencyForStatistics[$dependencies_projection_weight_property]) AS minWeight
7+
,max(dependencyForStatistics[$dependencies_projection_weight_property]) AS maxWeight
8+
,max(targetNodeForStatistics.maxDistanceFromSource) AS maxLevel
9+
WITH *, 1.0 / toFloat(maxWeight - minWeight) AS weightNormalizationFactor
10+
CALL gds.dag.longestPath.stream($dependencies_projection + '-cleaned')
11+
YIELD index, totalCost, path
12+
WITH *, toInteger(totalCost) AS distance
13+
ORDER BY distance DESC, index ASC
14+
UNWIND relationships(path) AS pathRelationship
15+
WITH *
16+
,startNode(pathRelationship) AS startNode
17+
,endNode(pathRelationship) AS endNode
18+
MATCH (startNode)-[dependency:DEPENDS_ON]->(endNode)
19+
WITH *, dependency[$dependencies_projection_weight_property] AS weight
20+
WITH *, toFloat(weight - minWeight) * weightNormalizationFactor AS normalizedWeight
21+
WITH *, round((normalizedWeight * 5) + 1, 2) AS penWidth
22+
WITH *, startNode.name + "\\n(level " + startNode.maxDistanceFromSource + "/" + maxLevel + ")" AS startNodeTitle
23+
WITH *, endNode.name + "\\n(level " + endNode.maxDistanceFromSource + "/" + maxLevel + ")" AS endNodeTitle
24+
WITH *, "[label=" + weight + "; penwidth=" + penWidth + "; ];" AS graphVizEdgeAttributes
25+
WITH *, "\"" + startNodeTitle + "\" -> \"" + endNodeTitle + "\" " + graphVizEdgeAttributes AS graphVizDotNotationLine
26+
RETURN graphVizDotNotationLine
27+
// Debugging
28+
// RETURN startNode.name AS startNodeName
29+
// ,endNode.name AS endNodeName
30+
// ,dependency[$dependencies_projection_weight_property] AS dependencyWeight
31+
// ,max(distance) AS partOfLongestPathLength
32+
// ,count(DISTINCT index) AS partOfLongestPathCounts
33+
// ,startNode.maxDistanceFromSource AS startNodeLevel
34+
// ,endNode.maxDistanceFromSource AS endNodeLevel
35+
LIMIT 100

‎cypher/Path_Finding/Set_Parameters.cypher‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
// Example on how to set the parameters for centrality in this case for Packages and PageRank
1+
// Example on how to set the parameters for path finding in this case for Packages and PageRank
22

33
:params {
44
"dependencies_projection": "package-path-finding",

‎scripts/reports/PathFindingVisualization.sh‎

Lines changed: 78 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,78 @@
1+
#!/usr/bin/env bash
2+
3+
# Executes selected "Path_Finding" Cypher queries for GraphViz visualization.
4+
# For example, the Java Artifact and TypeScript Module dependencies with the longest paths are visualized.
5+
# It requires an already running Neo4j graph database with already scanned and analyzed artifacts.
6+
# The reports (csv, dot and svg files) will be written into the sub directory reports/path-finding-visualization.
7+
8+
# Requires executeQueryFunctions.sh, projectionFunctions.sh, visualizeQueryResults.sh, cleanupAfterReportGeneration.sh
9+
10+
# Fail on any error ("-e" = exit on first error, "-o pipefail" exist on errors within piped commands)
11+
set -o errexit -o pipefail
12+
13+
# Overrideable Constants (defaults also defined in sub scripts)
14+
REPORTS_DIRECTORY=${REPORTS_DIRECTORY:-"reports"}
15+
16+
## Get this "scripts/reports" directory if not already set
17+
# Even if $BASH_SOURCE is made for Bourne-like shells it is also supported by others and therefore here the preferred solution.
18+
# CDPATH reduces the scope of the cd command to potentially prevent unintended directory changes.
19+
# This way non-standard tools like readlink aren't needed.
20+
REPORTS_SCRIPT_DIR=${REPORTS_SCRIPT_DIR:-$( CDPATH=. cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd -P )}
21+
echo "PathFindingVisualization: REPORTS_SCRIPT_DIR=${REPORTS_SCRIPT_DIR}"
22+
23+
# Get the "scripts" directory by taking the path of this script and going one directory up.
24+
SCRIPTS_DIR=${SCRIPTS_DIR:-"${REPORTS_SCRIPT_DIR}/.."} # Repository directory containing the shell scripts
25+
echo "PathFindingVisualization SCRIPTS_DIR=${SCRIPTS_DIR}"
26+
27+
# Get the "scripts/visualization" directory.
28+
VISUALIZATION_SCRIPTS_DIR=${VISUALIZATION_SCRIPTS_DIR:-"${SCRIPTS_DIR}/visualization"} # Repository directory containing the shell scripts for visualization
29+
echo "PathFindingVisualization VISUALIZATION_SCRIPTS_DIR=${VISUALIZATION_SCRIPTS_DIR}"
30+
31+
# Get the "cypher" directory by taking the path of this script and going two directory up and then to "cypher".
32+
CYPHER_DIR=${CYPHER_DIR:-"${REPORTS_SCRIPT_DIR}/../../cypher"}
33+
echo "PathFindingVisualization CYPHER_DIR=${CYPHER_DIR}"
34+
35+
PATH_FINDINGS_CYPHER_DIR="${CYPHER_DIR}/Path_Finding"
36+
37+
# Define functions to execute cypher queries from within a given file
38+
source "${SCRIPTS_DIR}/executeQueryFunctions.sh"
39+
40+
# Define functions to create and delete Graph Projections like "createDirectedDependencyProjection"
41+
source "${SCRIPTS_DIR}/projectionFunctions.sh"
42+
43+
# Create report directory
44+
REPORT_NAME="path-finding-visualization"
45+
FULL_REPORT_DIRECTORY="${REPORTS_DIRECTORY}/${REPORT_NAME}"
46+
mkdir -p "${FULL_REPORT_DIRECTORY}"
47+
48+
# Java Artifacts: Longest Paths Visualization
49+
ARTIFACT_PROJECTION="dependencies_projection=artifact-path-finding"
50+
ARTIFACT_NODE="dependencies_projection_node=Artifact"
51+
ARTIFACT_WEIGHT="dependencies_projection_weight_property=weight"
52+
if createDirectedDependencyProjection "${ARTIFACT_PROJECTION}" "${ARTIFACT_NODE}" "${ARTIFACT_WEIGHT}"; then
53+
reportName="${FULL_REPORT_DIRECTORY}/JavaArtifactLongestPathsIsolated"
54+
execute_cypher "${PATH_FINDINGS_CYPHER_DIR}/Path_Finding_6_Longest_paths_for_graphviz.cypher" "${ARTIFACT_PROJECTION}" "${ARTIFACT_NODE}" "${ARTIFACT_WEIGHT}" > "${reportName}.csv"
55+
source "${VISUALIZATION_SCRIPTS_DIR}/visualizeQueryResults.sh" "${reportName}.csv"
56+
57+
reportName="${FULL_REPORT_DIRECTORY}/JavaArtifactLongestPaths"
58+
execute_cypher "${PATH_FINDINGS_CYPHER_DIR}/Path_Finding_6_Longest_paths_contributors_for_graphviz.cypher" "${ARTIFACT_PROJECTION}" "${ARTIFACT_NODE}" "${ARTIFACT_WEIGHT}" > "${reportName}.csv"
59+
source "${VISUALIZATION_SCRIPTS_DIR}/visualizeQueryResults.sh" "${reportName}.csv"
60+
fi
61+
62+
# TypeScript Modules: Longest Paths Visualization
63+
MODULE_LANGUAGE="dependencies_projection_language=Typescript"
64+
MODULE_PROJECTION="dependencies_projection=typescript-module-path-finding"
65+
MODULE_NODE="dependencies_projection_node=Module"
66+
MODULE_WEIGHT="dependencies_projection_weight_property=lowCouplingElement25PercentWeight"
67+
if createDirectedDependencyProjection "${MODULE_LANGUAGE}" "${MODULE_PROJECTION}" "${MODULE_NODE}" "${MODULE_WEIGHT}"; then
68+
reportName="${FULL_REPORT_DIRECTORY}/TypeScriptModuleLongestPathsIsolated"
69+
execute_cypher "${PATH_FINDINGS_CYPHER_DIR}/Path_Finding_6_Longest_paths_for_graphviz.cypher" "${MODULE_PROJECTION}" "${MODULE_NODE}" "${MODULE_WEIGHT}" > "${reportName}.csv"
70+
source "${VISUALIZATION_SCRIPTS_DIR}/visualizeQueryResults.sh" "${reportName}.csv"
71+
72+
reportName="${FULL_REPORT_DIRECTORY}/TypeScriptModuleLongestPaths"
73+
execute_cypher "${PATH_FINDINGS_CYPHER_DIR}/Path_Finding_6_Longest_paths_contributors_for_graphviz.cypher" "${MODULE_PROJECTION}" "${MODULE_NODE}" "${MODULE_WEIGHT}" > "${reportName}.csv"
74+
source "${VISUALIZATION_SCRIPTS_DIR}/visualizeQueryResults.sh" "${reportName}.csv"
75+
fi
76+
77+
# Clean-up after report generation. Empty reports will be deleted.
78+
source "${SCRIPTS_DIR}/cleanupAfterReportGeneration.sh" "${FULL_REPORT_DIRECTORY}"

0 commit comments

Comments
 (0)