Outputs

Graphviz

This output leverages the GraphViz graph generation tool. You’ll need it to be installed before attempting to use it.

Gephi

This output generates a GDF file that can be used with Gephi.

JSON

This output writes the collected call graph as a single JSON document, so it can be consumed by other tooling without depending on a graph renderer.

pycallgraph json --output-file callgraph.json -- my_script.py

The document is versioned and contains nodes and edges:

{
  "version": 1,
  "nodes": [
    {"name": "my_module.main", "group": "my_module", "calls": 1,
     "time": 0.0012, "memory_in": 0, "memory_out": 0}
  ],
  "edges": [
    {"source": "my_module.main", "target": "my_module.helper",
     "calls": 2, "time": 0.0004}
  ]
}

Edge endpoints are named source/target (Gephi’s GDF output calls the same fields node1/node2).

Pickle

This output persists the completed trace so it can be loaded and rendered later without re-running the program:

pycallgraph pickle --output-file callgraph.pickle -- my_script.py

import pickle
with open('callgraph.pickle', 'rb') as handle:
    trace = pickle.load(handle)
for node in trace.nodes():
    print(node.name, node.calls.value)

A loaded trace can also drive any other output, which is the point of dumping:

from pycallgraph.output.graphviz import GraphvizOutput

output = GraphvizOutput()
output.output_file = 'rebuilt.png'
output.set_processor(trace)
output.done()

The dump stores the collected trace data plus the few settings needed to render it (grouping and memory flags). The trace filters, outputs and full command line configuration are not persisted.

Todo

Expand this section with screenshots and examples.