Helper Module for Deep Learning.
Source code for pynet.plotting.signal
# -*- coding: utf-8 -*-
##########################################################################
# NSAp - Copyright (C) CEA, 2019
# Distributed under the terms of the CeCILL-B license, as published by
# the CEA-CNRS-INRIA. Refer to the LICENSE file or to
# http://www.cecill.info/licences/Licence_CeCILL-B_V1-en.html
# for details.
##########################################################################
"""
Common functions to display signals.
"""
# Import
import numpy as np
import matplotlib.pyplot as plt
def _trim_axs(axs, size):
""" Little helper to massage the axs list to have correct length...
"""
axs = axs.flat
for ax in axs[size:]:
ax.remove()
return axs[:size]
[docs]def plot_history(history, title=None):
""" Plot an history.
Parameters
----------
history: pynet History
the history to be displayed.
title: str, default None
the figure title.
"""
nb_plots = len(history.metrics)
cols = 3
rows, rest = divmod(nb_plots, cols)
if rest > 0:
rows += 1
fig, axs = plt.subplots(rows, cols)
fig.title = title
axs = _trim_axs(axs, nb_plots)
for ax, case in zip(axs, history.metrics):
ax.set_title(case)
ax.plot(history[case][1])
Follow us
© 2019, pynet developers .
Inspired by AZMIND template.
Inspired by AZMIND template.