pylibstats is a small library that allows you to query some useful statistics for Python code-bases.
We currently report library imports, function calls and attributes.
You can install pylibstats with the following command
python -m pip install 'git+https://github.com/fmassa/python-lib-stats.git'
# (add --user if you don't have permission)
# Or, to install it from a local clone:
git clone https://github.com/fmassa/python-lib-stats.git
python -m pip install -e python-lib-stats
In order to find all uses (imports / function calls / attribute queries) from library <mylib> over codebase <path/to/python/repo>, run the following command:
pylibstats --local_dir <path/to/python/repo> --library_name <mylib> --absolute_count
You can also use pylibstats as a library.
import pylibstats summary = pylibstats.process_local_repository("path/to/repo", library_name="mylib")
where summary is a Dict[Dict[str, int]] with the following entries: import_count, call_count and access_count, each field being a dictionary with the mapping of name: count.
Looking for all torchvision occurrences on the DETR codebase yields:
===========================================================================
Imports
===========================================================================
| Count
---------------------------------------------------------------------------
torchvision : 4
torchvision.ops._new_empty_tensor : 1
torchvision.ops.misc._output_size : 1
torchvision.ops.boxes.box_area : 1
torchvision.transforms : 1
torchvision.transforms.functional : 1
torchvision.models._utils.IntermediateLayerGetter : 1
===========================================================================
Calls
===========================================================================
| Count
---------------------------------------------------------------------------
torchvision.__version__.split : 2
torchvision.ops.boxes.box_area : 2
torchvision.transforms.RandomCrop.get_params : 2
torchvision._is_tracing : 1
torchvision.ops.misc._output_size : 1
torchvision.ops._new_empty_tensor : 1
torchvision.ops.misc.interpolate : 1
torchvision.transforms.functional.crop : 1
torchvision.transforms.functional.hflip : 1
torchvision.transforms.functional.resize : 1
torchvision.transforms.functional.pad : 1
torchvision.transforms.functional.to_tensor : 1
torchvision.transforms.RandomErasing : 1
torchvision.transforms.functional.normalize : 1
torchvision.models._utils.IntermediateLayerGetter : 1
torchvision.models.{?} : 1
===========================================================================
Attrs
===========================================================================
| Count
---------------------------------------------------------------------------
torchvision.__version__.split : 2
torchvision.ops.boxes.box_area : 2
torchvision.datasets.CocoDetection : 2
torchvision.transforms.RandomCrop.get_params : 2
torchvision._is_tracing : 1
torchvision.ops.misc._output_size : 1
torchvision.ops._new_empty_tensor : 1
torchvision.ops.misc.interpolate : 1
torchvision.transforms.functional.crop : 1
torchvision.transforms.functional.hflip : 1
torchvision.transforms.functional.resize : 1
torchvision.transforms.functional.pad : 1
torchvision.transforms.functional.to_tensor : 1
torchvision.transforms.RandomErasing : 1
torchvision.transforms.functional.normalize : 1
torchvision.models._utils.IntermediateLayerGetter : 1
torchvision.models : 1