A static analysis library for computing graph representations of Python programs suitable for use with graph neural networks.

Last update: Dec 29, 2022

Related tags

Overview

python_graphs

This package is for computing graph representations of Python programs for machine learning applications. It includes the following modules:

control_flow For computing control flow graphs statically from Python programs.
data_flow For computing data flow analyses of Python programs.
program_graph For computing graphs statically to represent arbitrary Python programs or functions.
cyclomatic_complexity For computing the cyclomatic complexity of a Python function.

Installation

To install python_graphs with pip, run: pip install python_graphs.

To install python_graphs from source, run: python setup.py develop.

Common Tasks

Generate a control flow graph from a function fn:

from python_graphs import control_flow
graph = control_flow.get_control_flow_graph(fn)

Generate a program graph from a function fn:

from python_graphs import program_graph
graph = program_graph.get_program_graph(fn)

Compute the cyclomatic complexity of a function fn:

from python_graphs import control_flow
from python_graphs import cyclomatic_complexity
graph = control_flow.get_control_flow_graph(fn)
value = cyclomatic_complexity.cyclomatic_complexity(graph)

This is not an officially supported Google product.

Comments

Can you provide a quick start example?

Super cool project! Love the idea and think it has a lot of potential.

it would be awesome to have an examples/ directory containing some sample usages - maybe even just plotting the graphs with networkX and matplotlib.
question

opened by LukeWood 5
How do we solve the error when installing python-graphs?

Hello,

I encountered an error "fatal error: 'graphviz/cgraph.h' file not found" when trying to install python_graphs. How do I solve this issue, please? Thanks.
question

opened by fraolBatole 2
How to generate a Holistic Data Flow Graph for a given Function ?

@dbieber, Thanks for this awesome work.

Question

control_flow.get_control_flow_graph, returns a Control Flow Graph for a given Function Object. There is one data_flow class, Is there a way to generate a complete Data Flow Graph given a Function Object?

Thanks.

opened by reshinthadithyan 2
Rename fn to get_test_components to eliminate extra test from logs

The function test_components was being registered as an unsupported test, when in reality it was meant as a helper function for tests. Renaming resolves this.

opened by dbieber 0
get_start_control_flow_node, next_from_end, raise edges, and labels in branches
Adds get_start_control_flow_node to ControlFlowGraph

Adds next_from_end to ControlFlowNode

Uses labels (e.g. '' and '' strings) to indicate these special nodes

Support keyword only arguments without defaults

Add non-interrupting edges from raise statements

Bump version number
opened by dbieber 0
Separate branch kinds

Splits "branches" into branches, except_branches, and reraise_branches.

branches are you're usual branch decisions: ifs, fors, and whiles. except_branches are at "except E:" statements, with True indicating the exception matches and False indicating it does not reraise_branches are at the end of "finally:" blocks, with True indicating the path taken after finally if an error has been raised previously, and False indicating the path taken if there's nothing to reraise at the end of the finally.

opened by dbieber 0
Add module frame to catch raises in top-level code.

Add module frame to catch raises in top-level code. Also marks except expressions and finally blocks as branch points.

An "except A:"'s branch decision is whether the current exception matches A. At the end of a finally block, the branch decision is whether an exception is currently being raised.

This includes https://github.com/google-research/python-graphs/pull/3: Splits "branches" into branches, except_branches, and reraise_branches.

branches are your usual branch decisions: ifs, fors, and whiles. except_branches are at "except E:" statements, with True indicating the exception matches and False indicating it does not reraise_branches are at the end of "finally:" blocks, with True indicating the path taken after finally if an error has been raised previously, and False indicating the path taken if there's nothing to reraise at the end of the finally.

opened by dbieber 0
KeyError when trying to get program_graph
When I try to create a program graph, I encounter a KeyError. If I remove all the and and or expressions from the python file (buggy.py) the error does not occur.

This is how I use the library:

graph = program_graph.get_program_graph(code) program_graph_graphviz.render(graph, path='source.png')

where code is simply the code in the attached file buggy.py.txt.

I have also attached the log file log.txt.

buggy.py.txt

log.txt

More information: python 3.9.5 commit head=44c15b92197f374c3550353ff827997ef1c1d857 gast 0.5.3
opened by ppashakhanloo 1

Releases(v1.2.3)

v1.2.3(Oct 7, 2021)

get_start_control_flow_node, next_from_end, raise edges, and labels in branches (#6)

* Adds get_start_control_flow_node to ControlFlowGraph
* Adds next_from_end to ControlFlowNode
* Uses labels (e.g. '<exit>' and '<raise>' strings) to indicate these special nodes
* Support keyword only arguments without defaults
* Add non-interrupting edges from raise statements
* Bump version number

Source code(tar.gz)
Source code(zip)

v1.2.0(Oct 5, 2021)

Introduce get_branches API on control flow nodes. Previously the new branch types (except_branches and reraise_branches) were only accessible on basic blocks, not on individual nodes.
Source code(tar.gz)
Source code(zip)
v1.1.0(Oct 5, 2021)
Adds a module frame to catch raises in top-level code.

Also marks except expressions and finally blocks as branch points.

The branch kinds are: branches, except_branches, and reraise_branches.

branches are your usual branch decisions: ifs, fors, and whiles.

except_branches are at "except E:" statements, with True indicating the exception matches and False indicating it does not

reraise_branches are at the end of "finally:" blocks, with True indicating the path taken after finally if an error has been raised previously, and False indicating the path taken if there's nothing to reraise at the end of the finally.

Source code(tar.gz)
Source code(zip)
v1.0.1(May 7, 2021)

Includes __init__.py files to allow use of python_graphs library from Python 2.
Source code(tar.gz)
Source code(zip)
v1.0.0(Apr 12, 2021)
v1.0.0

Initial public release of the python_graphs library.

Core features:

control flow graph generation

data flow analyses

program graph construction

cyclomatic complexity calculation

a solid test suite for all the above

visualizations using graphviz for each of the graph representations

Source code(tar.gz)
Source code(zip)

Owner

Google Research

GitHub Repository

This repository contains the code used in the paper "Prompt-Based Multi-Modal Image Segmentation".

Prompt-Based Multi-Modal Image Segmentation This repository contains the code used in the paper "Prompt-Based Multi-Modal Image Segmentation". The sys

305 Dec 30, 2022

VOGUE: Try-On by StyleGAN Interpolation Optimization

VOGUE is a StyleGAN interpolation optimization algorithm for photo-realistic try-on. Top: shirt try-on automatically synthesized by our method in two different examples.

66 Dec 09, 2022

[CVPR2021 Oral] UP-DETR: Unsupervised Pre-training for Object Detection with Transformers

UP-DETR: Unsupervised Pre-training for Object Detection with Transformers This is the official PyTorch implementation and models for UP-DETR paper: @a

430 Dec 23, 2022

CV backbones including GhostNet, TinyNet and TNT, developed by Huawei Noah's Ark Lab.

CV Backbones including GhostNet, TinyNet, TNT (Transformer in Transformer) developed by Huawei Noah's Ark Lab. GhostNet Code TinyNet Code TNT Code Pyr

3k Jan 08, 2023

Implementation of "With a Little Help from my Temporal Context: Multimodal Egocentric Action Recognition, BMVC, 2021" in PyTorch

Multimodal Temporal Context Network (MTCN) This repository implements the model proposed in the paper: Evangelos Kazakos, Jaesung Huh, Arsha Nagrani,

13 Nov 24, 2022

code for paper "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning" by Zhongzheng Ren, Raymond A. Yeh, Alexander G. Schwing.

Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning Overview This code is for paper: Not All Unlabeled Data are Equa

22 Nov 23, 2022

This repository contains codes of ICCV2021 paper: SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation

SO-Pose This repository contains codes of ICCV2021 paper: SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation This paper is basically an

52 Nov 25, 2022

An Implementation of Transformer in Transformer in TensorFlow for image classification, attention inside local patches

Transformer-in-Transformer An Implementation of the Transformer in Transformer paper by Han et al. for image classification, attention inside local pa

40 Jul 25, 2022

A PyTorch Implementation of PGL-SUM from "Combining Global and Local Attention with Positional Encoding for Video Summarization", Proc. IEEE ISM 2021

PGL-SUM: Combining Global and Local Attention with Positional Encoding for Video Summarization PyTorch Implementation of PGL-SUM From "PGL-SUM: Combin

35 Dec 22, 2022

A static analysis library for computing graph representations of Python programs suitable for use with graph neural networks.

Related tags

Overview

python_graphs

Installation

Common Tasks

Comments

Question

Releases(v1.2.3)

v1.2.3(Oct 7, 2021)

v1.2.0(Oct 5, 2021)

v1.1.0(Oct 5, 2021)

v1.0.1(May 7, 2021)

v1.0.0(Apr 12, 2021)

v1.0.0

Core features:

Owner

Google Research

This repository contains the code used in the paper "Prompt-Based Multi-Modal Image Segmentation".

VOGUE: Try-On by StyleGAN Interpolation Optimization

[CVPR2021 Oral] UP-DETR: Unsupervised Pre-training for Object Detection with Transformers

CV backbones including GhostNet, TinyNet and TNT, developed by Huawei Noah's Ark Lab.

Implementation of "With a Little Help from my Temporal Context: Multimodal Egocentric Action Recognition, BMVC, 2021" in PyTorch

code for paper "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning" by Zhongzheng Ren*, Raymond A. Yeh*, Alexander G. Schwing.

This repository contains codes of ICCV2021 paper: SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation

An Implementation of Transformer in Transformer in TensorFlow for image classification, attention inside local patches

A PyTorch Implementation of PGL-SUM from "Combining Global and Local Attention with Positional Encoding for Video Summarization", Proc. IEEE ISM 2021

[ICRA 2022] An opensource framework for cooperative detection. Official implementation for OPV2V.

Citation Intent Classification in scientific papers using the Scicite dataset an Pytorch

Cosine Annealing With Warmup

PyTorch Implementation of "Non-Autoregressive Neural Machine Translation"

EMNLP 2021: Single-dataset Experts for Multi-dataset Question-Answering

Deep Learning Based Fasion Recommendation System for Ecommerce

This repository is a series of notebooks that show solutions for the projects at Dataquest.io.

Codes for AAAI22 paper "Learning to Solve Travelling Salesman Problem with Hardness-Adaptive Curriculum"

TensorFlow implementation of "A Simple Baseline for Bayesian Uncertainty in Deep Learning"

Code for ICCV2021 paper SPEC: Seeing People in the Wild with an Estimated Camera

AI-Fitness-Tracker - AI Fitness Tracker With Python

code for paper "Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning" by Zhongzheng Ren, Raymond A. Yeh, Alexander G. Schwing.