File post-processing#

Utilities for processing files generated during training/evaluation.


Exported memories#

This library provides an implementation for quickly loading exported memory files to inspect their contents in future post-processing steps. See the section Library utilities (skrl.utils module) for a real use case


Usage#

from skrl.utils import postprocessing


# assuming there is a directory called "memories" with Torch files in it
memory_iterator = postprocessing.MemoryFileIterator("memories/*.pt")
for filename, data in memory_iterator:
    filename    # str: basename of the current file
    data    # dict: keys are the names of the memory tensors in the file.
            # Tensor shapes are (memory size, number of envs, specific content size)

    # example of simple usage:
    # print the filenames of all memories and their tensor shapes
    print("\nfilename:", filename)
    print("  |-- states:", data['states'].shape)
    print("  |-- actions:", data['actions'].shape)
    print("  |-- rewards:", data['rewards'].shape)
    print("  |-- next_states:", data['next_states'].shape)
    print("  |-- dones:", data['dones'].shape)

API#

class skrl.utils.postprocessing.MemoryFileIterator(pathname: str)#

Bases: object

__init__(pathname: str) → None#

Python iterator for loading data from exported memories

The iterator will load the next memory file in the list of path names. The output of the iterator is a tuple of the filename and the memory data where the memory data is a dictionary of torch.Tensor (PyTorch), numpy.ndarray (NumPy) or lists (CSV) depending on the format and the keys of the dictionary are the names of the variables

Supported formats:

  • PyTorch (pt)

  • NumPy (npz)

  • Comma-separated values (csv)

Expected output shapes:

  • PyTorch: (memory_size, num_envs, data_size)

  • NumPy: (memory_size, num_envs, data_size)

  • Comma-separated values: (memory_size * num_envs, data_size)

Parameters:

pathname (str) – String containing a path specification for the exported memories. Python glob method is used to find all files matching the path specification

__iter__() → MemoryFileIterator#

Return self to make iterable

__next__() → Tuple[str, dict]#

Return next batch

Returns:

Tuple of file name and data

Return type:

tuple

_format_csv() → Tuple[str, dict]#

Load CSV file from file

Returns:

Tuple of file name and data

Return type:

tuple

_format_numpy() → Tuple[str, dict]#

Load numpy array from file

Returns:

Tuple of file name and data

Return type:

tuple

_format_torch() → Tuple[str, dict]#

Load PyTorch tensor from file

Returns:

Tuple of file name and data

Return type:

tuple


Tensorboard files#

This library provides an implementation for quickly loading Tensorboard files to inspect their contents in future post-processing steps. See the section Library utilities (skrl.utils module) for a real use case


Requirements#

This utility requires the TensorFlow package to be installed to load and parse Tensorboard files:

pip install tensorflow

Usage#

from skrl.utils import postprocessing


# assuming there is a directory called "runs" with experiments and Tensorboard files in it
tensorboard_iterator = postprocessing.TensorboardFileIterator("runs/*/events.out.tfevents.*", \
    tags=["Reward / Total reward (mean)"])
for dirname, data in tensorboard_iterator:
    dirname    # str: path of the directory (experiment name) containing the Tensorboard file
    data    # dict: keys are the tags, values are lists of [step, value] pairs

    # example of simple usage:
    # print the directory name and the value length for the "Reward / Total reward (mean)" tag
    print("\ndirname:", dirname)
    for tag, values in data.items():
        print("  |-- tag:", tag)
        print("  |   |-- value length:", len(values))

API#

class skrl.utils.postprocessing.TensorboardFileIterator(pathname: str, tags: str | List[str])#

Bases: object

__init__(pathname: str, tags: str | List[str]) → None#

Python iterator for loading data from Tensorboard files

The iterator will load the next Tensorboard file in the list of path names. The iterator’s output is a tuple of the directory name and the Tensorboard variables selected by the tags. The Tensorboard data is returned as a dictionary with the tag as the key and a list of steps and values as the value

Parameters:

  • pathname (str) –

    String containing a path specification for the Tensorboard files. Python glob method is used to find all files matching the path specification

  • tags (str or list of str) – String or list of strings containing the tags of the variables to load

__iter__() → TensorboardFileIterator#

Return self to make iterable

__next__() → Tuple[str, dict]#

Return next batch

Returns:

Tuple of directory name and data

Return type:

tuple