Logger

extensions.tensorboard.Logger.Logger(self, logdir='runs/', experiment=None)

Logger class to use tensorboard to visualize ANNarchy simulations. Requires the tensorboardX package (pip install tensorboardX).

The Logger class is a thin wrapper around tensorboardX.SummaryWriter, which you could also use directly. The doc is available at https://tensorboardx.readthedocs.io/. Tensorboard can read any logging data, as long as they are saved in the right format (tfevents), so it is not limited to tensorflow. TensorboardX has been developed to allow the use of tensorboard with pytorch.

The extension has to be imported explictly:

from ANNarchy.extensions.tensorboard import Logger

The Logger class has to be closed properly at the end of the script, so it is advised to use a context:

with Logger() as logger:
    logger.add_scalar("Accuracy", acc, trial)

You can also make sure to close it:

logger = Logger()
logger.add_scalar("Accuracy", acc, trial)
logger.close()

By default, the logs will be written in a subfolder of ./runs/ (which will be created in the current directory). The subfolder is a combination of the current datetime and of the hostname, e.g. ./runs/Apr22_12-11-22_machine. You can control these two elements by passing arguments to Logger():

with Logger(logdir="/tmp/annarchy", experiment="trial1"): # logs in /tmp/annarchy/trial1

The add_* methods allow you to log various structures, such as scalars, images, histograms, figures, etc.

A tag should be given to each plot. In the example above, the figure with the accuracy will be labelled “Accuracy” in tensorboard. You can also group plots together with tags such as “Global performance/Accuracy”, “Global performance/Error rate”, “Neural activity/Population 1”, etc.

After (or while) logging data within your simulation, run tensorboard in the terminal by specifying the log directory:

tensorboard --logdir runs

TensorboardX enqueues the data in memory before writing to disk. You can force flushing with:

logger.flush()

Parameters

Name Type Description Default
logdir str path (absolute or relative) to the logging directory. Subfolders will be created for each individual run. The default is “runs/” 'runs/'
experiment str name of the subfolder for the current run. By default, it is a combination of the current time and the hostname (e.g. Apr22_12-11-22_machine). If you reuse an experiment name, the data will be appended. None

Methods

Name Description
add_figure Logs a Matplotlib figure.
add_histogram Logs an histogram.
add_image Logs an image.
add_images Logs a set of images (e.g. receptive fields).
add_parameters Logs parameters of a simulation.
add_scalar Logs a single scalar value, e.g. a success rate at various stages of learning.
add_scalars Logs multiple scalar values to be displayed in the same figure, e.g. several metrics or neural activities.
close Closes the logger.
flush Forces the logged data to be flushed to disk.

add_figure

extensions.tensorboard.Logger.Logger.add_figure(
    tag,
    figure,
    step=None,
    close=True,
)

Logs a Matplotlib figure.

Example:

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        fig = plt.figure()
        plt.plot(pop.r)
        logger.add_figure("Activity", fig, trial)

Parameters

Name Type Description Default
tag str name of the image in tensorboard. required
figure list | np.ndarray a list or 1D numpy array of values. required
step int time index. None
close bool whether the logger will close the figure when done (default: True). True

add_histogram

extensions.tensorboard.Logger.Logger.add_histogram(tag, hist, step=None)

Logs an histogram.

Example:

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        weights= proj.w.flatten()
        logger.add_histogram("Weight distribution", weights, trial)

Parameters

Name Type Description Default
tag str name of the figure in tensorboard. required
hist list | np.ndarray a list or 1D numpy array of values. required
step int time index. None

add_image

extensions.tensorboard.Logger.Logger.add_image(
    tag,
    img,
    step=None,
    equalize=False,
)

Logs an image.

The image must be a numpy array of size (height, width) for monochrome images or (height, width, 3) for colored images. The values should either be integers between 0 and 255 or floats between 0 and 1. The parameter equalize forces the values to be between 0 and 1 by equalizing using the min/max values.

Example::

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        img = pop.r.reshape((10, 10))
        logger.add_image("Population / Firing rate", img, trial, equalize=True)

Parameters

Name Type Description Default
tag str name of the figure in tensorboard. required
img np.ndarray array for the image. required
step int time index. None
equalize bool rescales the pixels between 0 and 1 using the min and max values of the array. False

add_images

extensions.tensorboard.Logger.Logger.add_images(
    tag,
    img,
    step=None,
    equalize=False,
    equalize_per_image=False,
)

Logs a set of images (e.g. receptive fields).

The numpy array must be of size (number, height, width) for monochrome images or (number, height, width, 3) for colored images. The values should either be integers between 0 and 255 or floats between 0 and 1. The parameter equalize forces the values to be between 0 and 1 by equalizing using the min/max values.

Example:

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        weights= proj.w.reshape(100, 10, 10) # 100 post neurons, 10*10 pre neurons
        logger.add_images("Projection/Receptive fields", weights, trial, equalize=True)

Parameters

Name Type Description Default
tag str name of the figure in tensorboard. required
img np.array array for the images. required
step int time index. None
equalize bool rescales the pixels between 0 and 1 using the min and max values of the array. False
equalize_per_image bool whether the rescaling should be using the global min/max values of the array, or per image. Has no effect if equalize of False. False

add_parameters

extensions.tensorboard.Logger.Logger.add_parameters(params, metrics)

Logs parameters of a simulation.

This should be run only once per simulation, generally at the end. This allows to compare different runs of the same network using different parameter values and study how they influence the global output metrics, such as accuracy, error rate, reaction speed, etc.

Example:

with Logger() as logger:
    # ...
    logger.add_parameters({'learning_rate': lr, 'tau': tau}, {'accuracy': accuracy})

Parameters

Name Type Description Default
params dict dictionary of parameters. required
metrics dict dictionary of metrics. required

add_scalar

extensions.tensorboard.Logger.Logger.add_scalar(tag, value, step=None)

Logs a single scalar value, e.g. a success rate at various stages of learning.

Example:

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        accuracy = ...
        logger.add_scalar("Accuracy", accuracy, trial)

Parameters

Name Type Description Default
tag str name of the figure in tensorboard. required
value float value. required
step int time index. None

add_scalars

extensions.tensorboard.Logger.Logger.add_scalars(tag, value, step=None)

Logs multiple scalar values to be displayed in the same figure, e.g. several metrics or neural activities.

Example:

with Logger() as logger:
    for trial in range(100):
        simulate(1000.0)
        act1 = pop.r[0]
        act2 = pop.r[1]
        logger.add_scalars(
            "Accuracy", 
            {'First neuron': act1, 'Second neuron': act2}, 
            trial)

Parameters

Name Type Description Default
tag str name of the figure in tensorboard. required
value dict dictionary of values. required
step int time index. None

close

extensions.tensorboard.Logger.Logger.close()

Closes the logger.

flush

extensions.tensorboard.Logger.Logger.flush()

Forces the logged data to be flushed to disk.