def run(self):
print("Because Pylearn2 is under heavy development, we generally do "
"not advice using the `setup.py install` command. Please "
"consider using the `setup.py develop` command instead for the "
"following reasons:\n\n1. Using `setup.py install` creates a "
"copy of the Pylearn2 source code in your Python installation "
"path. In order to update Pylearn2 afterwards you will need to "
"rerun `setup.py install` (!). Simply using `git pull` to "
"update your local copy of Pylearn2 code will not suffice. \n\n"
"2. When using `sudo` to install Pylearn2, all files, "
"including the tutorials, will be copied to a directory owned "
"by root. Not only is running tutorials as root unsafe, it "
"also means that all Pylearn2-related environment variables "
"which were defined for the user will be unavailable.\n\n"
"Pressing enter will continue the installation of Pylearn2 in "
"`develop` mode instead. Note that this means that you need to "
"keep this folder with the Pylearn2 code in its current "
"location. If you know what you are doing, and are very sure "
"that you want to install Pylearn2 using the `install` "
"command instead, please type `install`.\n")
mode = None
while mode not in ['', 'install', 'develop', 'cancel']:
if mode is not None:
print("Please try again")
mode = input("Installation mode: [develop]/install/cancel: ")
if mode in ['', 'develop']:
self.distribution.run_command('develop')
if mode == 'install':
return install.run(self)
def get_choice(choice_to_explanation):
"""
.. todo::
WRITEME
Parameters
----------
choice_to_explanation : dict
Dictionary mapping possible user responses to strings describing
what that response will cause the script to do
Returns
-------
WRITEME
"""
d = choice_to_explanation
for key in d:
logger.info('\t{0}: {1}'.format(key, d[key]))
prompt = '/'.join(d.keys())+'? '
first = True
choice = ''
while first or choice not in d.keys():
if not first:
warnings.warn('unrecognized choice')
first = False
choice = input(prompt)
return choice
def run(self):
print("Because Pylearn2 is under heavy development, we generally do "
"not advice using the `setup.py install` command. Please "
"consider using the `setup.py develop` command instead for the "
"following reasons:\n\n1. Using `setup.py install` creates a "
"copy of the Pylearn2 source code in your Python installation "
"path. In order to update Pylearn2 afterwards you will need to "
"rerun `setup.py install` (!). Simply using `git pull` to "
"update your local copy of Pylearn2 code will not suffice. \n\n"
"2. When using `sudo` to install Pylearn2, all files, "
"including the tutorials, will be copied to a directory owned "
"by root. Not only is running tutorials as root unsafe, it "
"also means that all Pylearn2-related environment variables "
"which were defined for the user will be unavailable.\n\n"
"Pressing enter will continue the installation of Pylearn2 in "
"`develop` mode instead. Note that this means that you need to "
"keep this folder with the Pylearn2 code in its current "
"location. If you know what you are doing, and are very sure "
"that you want to install Pylearn2 using the `install` "
"command instead, please type `install`.\n")
mode = None
while mode not in ['', 'install', 'develop', 'cancel']:
if mode is not None:
print("Please try again")
mode = input("Installation mode: [develop]/install/cancel: ")
if mode in ['', 'develop']:
self.distribution.run_command('develop')
if mode == 'install':
return install.run(self)
def get_choice(choice_to_explanation):
"""
.. todo::
WRITEME
Parameters
----------
choice_to_explanation : dict
Dictionary mapping possible user responses to strings describing
what that response will cause the script to do
Returns
-------
WRITEME
"""
d = choice_to_explanation
for key in d:
logger.info('\t{0}: {1}'.format(key, d[key]))
prompt = '/'.join(d.keys()) + '? '
first = True
choice = ''
while first or choice not in d.keys():
if not first:
warnings.warn('unrecognized choice')
first = False
choice = input(prompt)
return choice
def remove_packages( packages_to_remove ):
"""
Uninstall packages, whether or not they
are found in the source.lst (so it can
remove datasets installed from file).
:param packages_to_remove: list of package names
:raises: IOErrors
"""
if packages_to_remove==[]:
raise RuntimeError("[rm] fatal: need packages names to remove.")
packages_really_to_remove=[]
for this_package in packages_to_remove:
if this_package in packages_sources:
#this_data_set_location=os.path.join(dataset_data_path,this_package)
# check if in the installed.lst
# then if directory actually exists
# then if you have rights to remove it
if this_package in installed_packages_list:
this_data_set_location=os.path.join( installed_packages_list[this_package].where,
installed_packages_list[this_package].name )
if os.path.exists(this_data_set_location):
if (file_access_rights(this_data_set_location,os.W_OK)):
# ok, you may have rights to delete it
packages_really_to_remove.append(this_package)
else:
logger.warning("[rm] insufficient rights "
"to remove '{0}'".format(this_package))
else:
logger.warning("[rm] package '{0}' found in config file "
"but not installed".format(this_package))
else:
logger.warning("[rm] package '{0}' "
"not installed".format(this_package))
else:
logger.warning("[rm] unknown package '{0}'".format(this_package))
if packages_really_to_remove!=[]:
logger.info("[rm] the following packages will be removed permanently:")
packages = []
for this_package in packages_really_to_remove:
packages.append(this_package)
logger.info(' '.join(packages))
r = input("Proceed? [yes/N] ")
if r=='y' or r=='yes':
for this_package in packages_really_to_remove:
remove_package( installed_packages_list[this_package], dataset_data_path )
else:
logger.info("[up] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to remove, filenames where bad.
pass
def install_packages_from_file(packages_to_install):
"""
(Force)Installs packages from files, but does
not update installed.lst files.
caveat: not as tested as everything else.
:param packages_to_install: list of files to install
:raises: IOErrors
"""
if packages_to_install == []:
raise RuntimeError("[in] fatal: need packages names to install.")
packages_really_to_install = []
for this_package in packages_to_install:
if os.path.exists(this_package):
packages_really_to_install.append(this_package)
else:
logger.warning("[in] package '{0}' not found".format(this_package))
if packages_really_to_install != []:
logger.info("[in] The following package(s) will be installed:")
packages = []
for this_package in packages_really_to_install:
packages.append(corename(this_package))
logger.info(' '.join(packages))
r = input("Proceed? [yes/N] ")
if r == 'y' or r == 'yes':
for this_package in packages_really_to_install:
#install_upgrade( this_package, upgrade=False, progress_hook=hook )
if os.path.exists(dataset_data_path + corename(this_package)):
r = input(
"[in] '%s' already installed, overwrite? [yes/N] " %
corename(this_package))
if r != 'y' and r != 'yes':
logger.info("[in] skipping package "
"'{0}'".format(corename(this_package)))
continue
install_package(corename(this_package), this_package,
dataset_data_path)
#update_installed_list("i",(make a package object here),dataset_data_path)
else:
logger.info("[in] Taking '{0}' for no, so there.".format(r))
def install_packages_from_file( packages_to_install ):
"""
(Force)Installs packages from files, but does
not update installed.lst files.
caveat: not as tested as everything else.
:param packages_to_install: list of files to install
:raises: IOErrors
"""
if packages_to_install==[]:
raise RuntimeError("[in] fatal: need packages names to install.")
packages_really_to_install=[]
for this_package in packages_to_install:
if os.path.exists(this_package):
packages_really_to_install.append(this_package)
else:
logger.warning("[in] package '{0}' not found".format(this_package))
if packages_really_to_install!=[]:
logger.info("[in] The following package(s) will be installed:")
packages = []
for this_package in packages_really_to_install:
packages.append(corename(this_package))
logger.info(' '.join(packages))
r = input("Proceed? [yes/N] ")
if r=='y' or r=='yes':
for this_package in packages_really_to_install:
#install_upgrade( this_package, upgrade=False, progress_hook=hook )
if os.path.exists(dataset_data_path+corename(this_package)):
r = input("[in] '%s' already installed, overwrite? [yes/N] " % corename(this_package))
if r!='y' and r!='yes':
logger.info("[in] skipping package "
"'{0}'".format(corename(this_package)))
continue
install_package( corename(this_package), this_package, dataset_data_path)
#update_installed_list("i",(make a package object here),dataset_data_path)
else:
logger.info("[in] Taking '{0}' for no, so there.".format(r))
def show_reconstructions(m, model_path):
"""
Show reconstructions of a given DBM model.
Parameters
----------
m: int
rows * cols
model_path: str
Path of the model.
"""
model = load_model(model_path, m)
x = input('use test set? (y/n) ')
dataset = load_dataset(model.dataset_yaml_src, x)
vis_batch = dataset.get_batch_topo(m)
pv = init_viewer(dataset, rows, cols, vis_batch)
batch = model.visible_layer.space.make_theano_batch()
reconstruction = model.reconstruct(batch)
recons_func = function([batch], reconstruction)
if hasattr(model.visible_layer, 'beta'):
beta = model.visible_layer.beta.get_value()
print('beta: ', (beta.min(), beta.mean(), beta.max()))
while True:
update_viewer(dataset, batch, rows, cols, pv, recons_func, vis_batch)
pv.show()
print('Displaying reconstructions. (q to quit, ENTER = show more)')
while True:
x = input()
if x == 'q':
quit()
if x == '':
x = 1
break
else:
print('Invalid input, try again')
vis_batch = dataset.get_batch_topo(m)
def show_reconstructions(m, model_path):
"""
Show reconstructions of a given DBM model.
Parameters
----------
m: int
rows * cols
model_path: str
Path of the model.
"""
model = load_model(model_path, m)
x = input('use test set? (y/n) ')
dataset = load_dataset(model.dataset_yaml_src, x)
vis_batch = dataset.get_batch_topo(m)
pv = init_viewer(dataset, rows, cols, vis_batch)
batch = model.visible_layer.space.make_theano_batch()
reconstruction = model.reconstruct(batch)
recons_func = function([batch], reconstruction)
if hasattr(model.visible_layer, 'beta'):
beta = model.visible_layer.beta.get_value()
print('beta: ', (beta.min(), beta.mean(), beta.max()))
while True:
update_viewer(dataset, batch, rows, cols, pv, recons_func, vis_batch)
pv.show()
print('Displaying reconstructions. (q to quit, ENTER = show more)')
while True:
x = input()
if x == 'q':
quit()
if x == '':
x = 1
break
else:
print('Invalid input, try again')
vis_batch = dataset.get_batch_topo(m)
def install_packages(packages_to_install, force_install=False, hook=None):
"""
Installs the packages, possibly forcing installs.
:param packages_to_install: list of package names
:param force_install: if True, re-installs even if installed.
:param hook: download progress hook
:raises: IOErrors
"""
if packages_to_install == []:
raise RuntimeError("[in] fatal: need packages names to install.")
if force_install:
logger.warning("[in] using the force")
packages_really_to_install = []
for this_package in packages_to_install:
if this_package in packages_sources:
if force_install or not this_package in installed_packages_list:
packages_really_to_install.append(this_package)
else:
logger.warning("[in] package '{0}' "
"is already installed".format(this_package))
else:
logger.warning("[in] unknown package '{0}'".format(this_package))
if packages_really_to_install != []:
logger.info("[in] The following package(s) will be installed:")
for this_package in packages_really_to_install:
readable_size = packages_sources[this_package].readable_size
logger.info("{0} ({1})".format(this_package, readable_size))
r = input("Proceed? [yes/N] ")
if r == 'y' or r == 'yes':
for this_package in packages_really_to_install:
install_upgrade(packages_sources[this_package],
upgrade=False,
progress_hook=hook)
else:
logger.info("[in] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to upgrade,
# move along.
pass
def install_packages( packages_to_install, force_install=False, hook=None ):
"""
Installs the packages, possibly forcing installs.
:param packages_to_install: list of package names
:param force_install: if True, re-installs even if installed.
:param hook: download progress hook
:raises: IOErrors
"""
if packages_to_install==[]:
raise RuntimeError("[in] fatal: need packages names to install.")
if force_install:
logger.warning("[in] using the force")
packages_really_to_install=[]
for this_package in packages_to_install:
if this_package in packages_sources:
if force_install or not this_package in installed_packages_list:
packages_really_to_install.append(this_package)
else:
logger.warning("[in] package '{0}' "
"is already installed".format(this_package))
else:
logger.warning("[in] unknown package '{0}'".format(this_package))
if packages_really_to_install!=[]:
logger.info("[in] The following package(s) will be installed:")
for this_package in packages_really_to_install:
readable_size = packages_sources[this_package].readable_size
logger.info("{0} ({1})".format(this_package, readable_size))
r = input("Proceed? [yes/N] ")
if r=='y' or r=='yes':
for this_package in packages_really_to_install:
install_upgrade( packages_sources[this_package], upgrade=False, progress_hook=hook )
else:
logger.info("[in] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to upgrade,
# move along.
pass
def create_archive(source, archive_name):
if os.path.exists(archive_name):
r = input("'%s' exists, overwrite? [yes/N] " % archive_name)
if (r != "y") and (r != "yes"):
logger.info("taking '{0}' for no, so there.".format(r))
#bail out
return
try:
tar = tarfile.open(archive_name, mode="w:bz2")
except Exception as e:
logger.exception(e)
return
else:
for root, dirs, files in os.walk(source):
for filename in files:
this_file = os.path.join(root, filename)
logger.info("adding '{0}'".format(this_file))
tar.add(this_file)
tar.close()
def create_archive( source, archive_name ):
if os.path.exists(archive_name):
r = input("'%s' exists, overwrite? [yes/N] " % archive_name)
if (r!="y") and (r!="yes"):
logger.info("taking '{0}' for no, so there.".format(r))
#bail out
return
try:
tar=tarfile.open(archive_name,mode="w:bz2")
except Exception as e:
logger.exception(e)
return
else:
for root, dirs, files in os.walk(source):
for filename in files:
this_file = os.path.join(root,filename)
logger.info("adding '{0}'".format(this_file))
tar.add(this_file)
tar.close()
cols = 10
m = rows * cols
_, model_path = sys.argv
print('Loading model...')
model = serial.load(model_path)
model.set_batch_size(m)
dataset_yaml_src = model.dataset_yaml_src
print('Loading data...')
dataset = yaml_parse.load(dataset_yaml_src)
x = input('use test set? (y/n) ')
if x == 'y':
dataset = dataset.get_test_set()
else:
assert x == 'n'
vis_batch = dataset.get_batch_topo(m)
_, patch_rows, patch_cols, channels = vis_batch.shape
assert _ == m
mapback = hasattr(dataset, 'mapback_for_viewer')
actual_cols = 2 * cols * (1 + mapback) * (1 + (channels == 2))
pv = PatchViewer((1, 2), (r, c), is_color=False)
i = 0
while True:
patch = topo[i, :, :, :]
patch = patch / np.abs(patch).max()
pv.add_patch(patch[:, :, 1], rescale=False)
pv.add_patch(patch[:, :, 0], rescale=False)
pv.show()
print(dataset.y[i])
choices = {'g': 'goto image', 'q': 'quit'}
if i + 1 < b:
choices['n'] = 'next image'
choice = get_choice(choices)
if choice == 'q':
quit()
if choice == 'n':
i += 1
if choice == 'g':
i = int(input('index: '))
def upgrade_packages(packages_to_upgrade, hook=None ):
"""
Upgrades packages.
If no packages are supplied, it will perform
an "update-all" operation, finding all packages
that are out of date.
If packages names are supplied, only those
are checked for upgrade (and upgraded if out
of date)
:param packages_to_upgrade: list of package names.
:raises: IOErrors (from downloads/rights)
"""
# get names only
if packages_to_upgrade==[]:
packages_to_upgrade=installed_packages_list.keys() # all installed!
all_packages=True
else:
all_packages=False
# check what packages are in the list,
# and really to be upgraded.
#
packages_really_to_upgrade=[]
for this_package in packages_to_upgrade:
if this_package in installed_packages_list:
# check if there's a date
installed_date=installed_packages_list[this_package].timestamp
if this_package in packages_sources:
repo_date=packages_sources[this_package].timestamp
if installed_date < repo_date:
# ok, there's a newer version
logger.info(this_package)
packages_really_to_upgrade.append(this_package)
else:
# no newer version, nothing to update
pass
else:
logger.warning("[up] '{0}' is unknown "
"(installed from file?).".format(this_package))
else:
# not installed?
if not all_packages:
logger.warning("[up] '{0}' is not installed, "
"cannot upgrade.".format(this_package))
pass
# once we have determined which packages
# are to be updated, we show them to the
# user for him to confirm
#
if packages_really_to_upgrade!=[]:
logger.info("[up] the following package(s) will be upgraded:")
for this_package in packages_really_to_upgrade:
readable_size = packages_sources[this_package].readable_size
logger.info("{0} ({1})".format(this_package, readable_size))
r = input("Proceed? [yes/N] ")
if r=='y' or r=='yes':
for this_package in packages_really_to_upgrade:
install_upgrade( packages_sources[this_package], upgrade=True, progress_hook=hook )
else:
logger.info("[up] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to upgrade,
# move along.
pass
def get_weights_report(model_path=None,
model=None,
rescale='individual',
border=False,
norm_sort=False,
dataset=None):
"""
Returns a PatchViewer displaying a grid of filter weights
Parameters
----------
model_path : str
Filepath of the model to make the report on.
rescale : str
A string specifying how to rescale the filter images:
- 'individual' (default) : scale each filter so that it
uses as much as possible of the dynamic range
of the display under the constraint that 0
is gray and no value gets clipped
- 'global' : scale the whole ensemble of weights
- 'none' : don't rescale
dataset : pylearn2.datasets.dataset.Dataset
Dataset object to do view conversion for displaying the weights. If
not provided one will be loaded from the model's dataset_yaml_src.
Returns
-------
WRITEME
"""
if model is None:
logger.info('making weights report')
logger.info('loading model')
model = serial.load(model_path)
logger.info('loading done')
else:
assert model_path is None
assert model is not None
if rescale == 'none':
global_rescale = False
patch_rescale = False
elif rescale == 'global':
global_rescale = True
patch_rescale = False
elif rescale == 'individual':
global_rescale = False
patch_rescale = True
else:
raise ValueError('rescale=' + rescale +
", must be 'none', 'global', or 'individual'")
if isinstance(model, dict):
#assume this was a saved matlab dictionary
del model['__version__']
del model['__header__']
del model['__globals__']
keys = [key for key in model \
if hasattr(model[key], 'ndim') and model[key].ndim == 2]
if len(keys) > 2:
key = None
while key not in keys:
logger.info('Which is the weights?')
for key in keys:
logger.info('\t{0}'.format(key))
key = input()
else:
key, = keys
weights = model[key]
norms = np.sqrt(np.square(weights).sum(axis=1))
logger.info('min norm: {0}'.format(norms.min()))
logger.info('mean norm: {0}'.format(norms.mean()))
logger.info('max norm: {0}'.format(norms.max()))
return patch_viewer.make_viewer(weights,
is_color=weights.shape[1] % 3 == 0)
weights_view = None
W = None
try:
weights_view = model.get_weights_topo()
h = weights_view.shape[0]
except NotImplementedError:
if dataset is None:
logger.info('loading dataset...')
control.push_load_data(False)
dataset = yaml_parse.load(model.dataset_yaml_src)
control.pop_load_data()
logger.info('...done')
try:
W = model.get_weights()
except AttributeError as e:
reraise_as(AttributeError("""
Encountered an AttributeError while trying to call get_weights on a model.
This probably means you need to implement get_weights for this model class,
but look at the original exception to be sure.
If this is an older model class, it may have weights stored as weightsShared,
etc.
Original exception: """+str(e)))
if W is None and weights_view is None:
raise ValueError("model doesn't support any weights interfaces")
if weights_view is None:
weights_format = model.get_weights_format()
assert hasattr(weights_format,'__iter__')
assert len(weights_format) == 2
assert weights_format[0] in ['v','h']
assert weights_format[1] in ['v','h']
assert weights_format[0] != weights_format[1]
if weights_format[0] == 'v':
W = W.T
h = W.shape[0]
if norm_sort:
norms = np.sqrt(1e-8+np.square(W).sum(axis=1))
norm_prop = norms / norms.max()
weights_view = dataset.get_weights_view(W)
assert weights_view.shape[0] == h
try:
hr, hc = model.get_weights_view_shape()
except NotImplementedError:
hr = int(np.ceil(np.sqrt(h)))
hc = hr
if 'hidShape' in dir(model):
hr, hc = model.hidShape
pv = patch_viewer.PatchViewer(grid_shape=(hr, hc),
patch_shape=weights_view.shape[1:3],
is_color = weights_view.shape[-1] == 3)
if global_rescale:
weights_view /= np.abs(weights_view).max()
if norm_sort:
logger.info('sorting weights by decreasing norm')
idx = sorted( range(h), key=lambda l : - norm_prop[l] )
else:
idx = range(h)
if border:
act = 0
else:
act = None
for i in range(0,h):
patch = weights_view[idx[i],...]
pv.add_patch(patch, rescale=patch_rescale, activation=act)
abs_weights = np.abs(weights_view)
logger.info('smallest enc weight magnitude: {0}'.format(abs_weights.min()))
logger.info('mean enc weight magnitude: {0}'.format(abs_weights.mean()))
logger.info('max enc weight magnitude: {0}'.format(abs_weights.max()))
if W is not None:
norms = np.sqrt(np.square(W).sum(axis=1))
assert norms.shape == (h,)
logger.info('min norm: {0}'.format(norms.min()))
logger.info('mean norm: {0}'.format(norms.mean()))
logger.info('max norm: {0}'.format(norms.max()))
return pv
def main():
"""
.. todo::
WRITEME
"""
parser = argparse.ArgumentParser()
parser.add_argument("--out")
parser.add_argument("model_paths", nargs='+')
parser.add_argument("--yrange",
help='The y-range to be used for plotting, e.g. 0:1')
options = parser.parse_args()
model_paths = options.model_paths
if options.out is not None:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
print('generating names...')
model_names = [
model_path.replace('.pkl', '!') for model_path in model_paths
]
model_names = unique_substrings(model_names, min_size=10)
model_names = [model_name.replace('!', '') for model_name in model_names]
print('...done')
for i, arg in enumerate(model_paths):
try:
model = serial.load(arg)
except Exception:
if arg.endswith('.yaml'):
print(sys.stderr,
arg + " is a yaml config file," +
"you need to load a trained model.",
file=sys.stderr)
quit(-1)
raise
this_model_channels = model.monitor.channels
if len(sys.argv) > 2:
postfix = ":" + model_names[i]
else:
postfix = ""
for channel in this_model_channels:
channels[channel + postfix] = this_model_channels[channel]
del model
gc.collect()
while True:
# Make a list of short codes for each channel so user can specify them
# easily
tag_generator = _TagGenerator()
codebook = {}
sorted_codes = []
for channel_name in sorted(channels,
key=number_aware_alphabetical_key):
code = tag_generator.get_tag()
codebook[code] = channel_name
codebook['<' + channel_name + '>'] = channel_name
sorted_codes.append(code)
x_axis = 'example'
print('set x_axis to example')
if len(channels.values()) == 0:
print("there are no channels to plot")
break
# If there is more than one channel in the monitor ask which ones to
# plot
prompt = len(channels.values()) > 1
if prompt:
# Display the codebook
for code in sorted_codes:
print(code + '. ' + codebook[code])
print()
print("Put e, b, s or h in the list somewhere to plot " +
"epochs, batches, seconds, or hours, respectively.")
response = input('Enter a list of channels to plot ' + \
'(example: A, C,F-G, h, <test_err>) or q to quit' + \
' or o for options: ')
if response == 'o':
print('1: smooth all channels')
print('any other response: do nothing, go back to plotting')
response = input('Enter your choice: ')
if response == '1':
for channel in channels.values():
k = 5
new_val_record = []
for i in xrange(len(channel.val_record)):
new_val = 0.
count = 0.
for j in xrange(max(0, i - k), i + 1):
new_val += channel.val_record[j]
count += 1.
new_val_record.append(new_val / count)
channel.val_record = new_val_record
continue
if response == 'q':
break
#Remove spaces
response = response.replace(' ', '')
#Split into list
codes = response.split(',')
final_codes = set([])
for code in codes:
if code == 'e':
x_axis = 'epoch'
continue
elif code == 'b':
x_axis = 'batche'
elif code == 's':
x_axis = 'second'
elif code == 'h':
x_axis = 'hour'
elif code.startswith('<'):
assert code.endswith('>')
final_codes.add(code)
elif code.find('-') != -1:
#The current list element is a range of codes
rng = code.split('-')
if len(rng) != 2:
print("Input not understood: " + code)
quit(-1)
found = False
for i in xrange(len(sorted_codes)):
if sorted_codes[i] == rng[0]:
found = True
break
if not found:
print("Invalid code: " + rng[0])
quit(-1)
found = False
for j in xrange(i, len(sorted_codes)):
if sorted_codes[j] == rng[1]:
found = True
break
if not found:
print("Invalid code: " + rng[1])
quit(-1)
final_codes = final_codes.union(set(sorted_codes[i:j + 1]))
else:
#The current list element is just a single code
final_codes = final_codes.union(set([code]))
# end for code in codes
else:
final_codes, = set(codebook.keys())
colors = ['b', 'g', 'r', 'c', 'm', 'y', 'k']
styles = list(colors)
styles += [color + '--' for color in colors]
styles += [color + ':' for color in colors]
fig = plt.figure()
ax = plt.subplot(1, 1, 1)
# plot the requested channels
for idx, code in enumerate(sorted(final_codes)):
channel_name = codebook[code]
channel = channels[channel_name]
y = np.asarray(channel.val_record)
if contains_nan(y):
print(channel_name + ' contains NaNs')
if contains_inf(y):
print(channel_name + 'contains infinite values')
if x_axis == 'example':
x = np.asarray(channel.example_record)
elif x_axis == 'batche':
x = np.asarray(channel.batch_record)
elif x_axis == 'epoch':
try:
x = np.asarray(channel.epoch_record)
except AttributeError:
# older saved monitors won't have epoch_record
x = np.arange(len(channel.batch_record))
elif x_axis == 'second':
x = np.asarray(channel.time_record)
elif x_axis == 'hour':
x = np.asarray(channel.time_record) / 3600.
else:
assert False
ax.plot(
x,
y,
styles[idx % len(styles)],
marker='.', # add point margers to lines
label=channel_name)
plt.xlabel('# ' + x_axis + 's')
ax.ticklabel_format(scilimits=(-3, 3), axis='both')
handles, labels = ax.get_legend_handles_labels()
lgd = ax.legend(handles,
labels,
loc='upper center',
bbox_to_anchor=(0.5, -0.1))
# 0.046 is the size of 1 legend box
fig.subplots_adjust(bottom=0.11 + 0.046 * len(final_codes))
if (options.yrange is not None):
ymin, ymax = map(float, options.yrange.split(':'))
plt.ylim(ymin, ymax)
if options.out is None:
plt.show()
else:
plt.savefig(options.out)
if not prompt:
break
def main():
"""
.. todo::
WRITEME
"""
parser = argparse.ArgumentParser()
parser.add_argument("--out")
parser.add_argument("model_paths", nargs='+')
options = parser.parse_args()
model_paths = options.model_paths
if options.out is not None:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
print('generating names...')
model_names = [model_path.replace('.pkl', '!') for model_path in
model_paths]
model_names = unique_substrings(model_names, min_size=10)
model_names = [model_name.replace('!','') for model_name in
model_names]
print('...done')
for i, arg in enumerate(model_paths):
try:
model = serial.load(arg)
except Exception:
if arg.endswith('.yaml'):
print(sys.stderr, arg + " is a yaml config file," +
"you need to load a trained model.", file=sys.stderr)
quit(-1)
raise
this_model_channels = model.monitor.channels
if len(sys.argv) > 2:
postfix = ":" + model_names[i]
else:
postfix = ""
for channel in this_model_channels:
channels[channel+postfix] = this_model_channels[channel]
del model
gc.collect()
while True:
# Make a list of short codes for each channel so user can specify them
# easily
tag_generator = _TagGenerator()
codebook = {}
sorted_codes = []
for channel_name in sorted(channels,
key = number_aware_alphabetical_key):
code = tag_generator.get_tag()
codebook[code] = channel_name
codebook['<'+channel_name+'>'] = channel_name
sorted_codes.append(code)
x_axis = 'example'
print('set x_axis to example')
if len(channels.values()) == 0:
print("there are no channels to plot")
break
# If there is more than one channel in the monitor ask which ones to
# plot
prompt = len(channels.values()) > 1
if prompt:
# Display the codebook
for code in sorted_codes:
print(code + '. ' + codebook[code])
print()
print("Put e, b, s or h in the list somewhere to plot " +
"epochs, batches, seconds, or hours, respectively.")
response = input('Enter a list of channels to plot ' + \
'(example: A, C,F-G, h, <test_err>) or q to quit' + \
' or o for options: ')
if response == 'o':
print('1: smooth all channels')
print('any other response: do nothing, go back to plotting')
response = input('Enter your choice: ')
if response == '1':
for channel in channels.values():
k = 5
new_val_record = []
for i in xrange(len(channel.val_record)):
new_val = 0.
count = 0.
for j in xrange(max(0, i-k), i+1):
new_val += channel.val_record[j]
count += 1.
new_val_record.append(new_val / count)
channel.val_record = new_val_record
continue
if response == 'q':
break
#Remove spaces
response = response.replace(' ','')
#Split into list
codes = response.split(',')
final_codes = set([])
for code in codes:
if code == 'e':
x_axis = 'epoch'
continue
elif code == 'b':
x_axis = 'batche'
elif code == 's':
x_axis = 'second'
elif code == 'h':
x_axis = 'hour'
elif code.startswith('<'):
assert code.endswith('>')
final_codes.add(code)
elif code.find('-') != -1:
#The current list element is a range of codes
rng = code.split('-')
if len(rng) != 2:
print("Input not understood: "+code)
quit(-1)
found = False
for i in xrange(len(sorted_codes)):
if sorted_codes[i] == rng[0]:
found = True
break
if not found:
print("Invalid code: "+rng[0])
quit(-1)
found = False
for j in xrange(i,len(sorted_codes)):
if sorted_codes[j] == rng[1]:
found = True
break
if not found:
print("Invalid code: "+rng[1])
quit(-1)
final_codes = final_codes.union(set(sorted_codes[i:j+1]))
else:
#The current list element is just a single code
final_codes = final_codes.union(set([code]))
# end for code in codes
else:
final_codes ,= set(codebook.keys())
colors = ['b', 'g', 'r', 'c', 'm', 'y', 'k']
styles = list(colors)
styles += [color+'--' for color in colors]
styles += [color+':' for color in colors]
fig = plt.figure()
ax = plt.subplot(1,1,1)
# plot the requested channels
for idx, code in enumerate(sorted(final_codes)):
channel_name= codebook[code]
channel = channels[channel_name]
y = np.asarray(channel.val_record)
if contains_nan(y):
print(channel_name + ' contains NaNs')
if contains_inf(y):
print(channel_name + 'contains infinite values')
if x_axis == 'example':
x = np.asarray(channel.example_record)
elif x_axis == 'batche':
x = np.asarray(channel.batch_record)
elif x_axis == 'epoch':
try:
x = np.asarray(channel.epoch_record)
except AttributeError:
# older saved monitors won't have epoch_record
x = np.arange(len(channel.batch_record))
elif x_axis == 'second':
x = np.asarray(channel.time_record)
elif x_axis == 'hour':
x = np.asarray(channel.time_record) / 3600.
else:
assert False
ax.plot( x,
y,
styles[idx % len(styles)],
marker = '.', # add point margers to lines
label = channel_name)
plt.xlabel('# '+x_axis+'s')
ax.ticklabel_format( scilimits = (-3,3), axis = 'both')
handles, labels = ax.get_legend_handles_labels()
lgd = ax.legend(handles, labels, loc='upper center',
bbox_to_anchor=(0.5,-0.1))
# 0.046 is the size of 1 legend box
fig.subplots_adjust(bottom=0.11 + 0.046 * len(final_codes))
if options.out is None:
plt.show()
else:
plt.savefig(options.out)
if not prompt:
break
def show_samples(m, model_path):
"""
Show samples given a DBM model.
Parameters
----------
m: int
rows * cols
model_path: str
Path of the model.
"""
model = load_model(model_path, m)
print('Loading data (used for setting up visualization '
'and seeding gibbs chain) ...')
dataset_yaml_src = model.dataset_yaml_src
dataset = yaml_parse.load(dataset_yaml_src)
pv = init_viewer(dataset, rows, cols)
if hasattr(model.visible_layer, 'beta'):
beta = model.visible_layer.beta.get_value()
print('beta: ', (beta.min(), beta.mean(), beta.max()))
print('showing seed data...')
vis_batch = dataset.get_batch_topo(m)
update_viewer(dataset, pv, vis_batch, rows, cols)
pv.show()
print('How many Gibbs steps should I run with the seed data clamped?'
'(negative = ignore seed data)')
x = int(input())
# Make shared variables representing the sampling state of the model
layer_to_state = model.make_layer_to_state(m)
# Seed the sampling with the data batch
vis_sample = layer_to_state[model.visible_layer]
validate_all_samples(model, layer_to_state)
if x >= 0:
if vis_sample.ndim == 4:
vis_sample.set_value(vis_batch)
else:
design_matrix = dataset.get_design_matrix(vis_batch)
vis_sample.set_value(design_matrix)
validate_all_samples(model, layer_to_state)
sample_func = get_sample_func(model, layer_to_state, x)
while True:
print('Displaying samples. '
'How many steps to take next? (q to quit, ENTER=1)')
while True:
x = input()
if x == 'q':
quit()
if x == '':
x = 1
break
else:
try:
x = int(x)
break
except ValueError:
print('Invalid input, try again')
for i in xrange(x):
print(i)
sample_func()
validate_all_samples(model, layer_to_state)
vis_batch = vis_sample.get_value()
update_viewer(dataset, pv, vis_batch, rows, cols)
pv.show()
if 'Softmax' in str(type(model.hidden_layers[-1])):
state = layer_to_state[model.hidden_layers[-1]]
value = state.get_value()
y = np.argmax(value, axis=1)
assert y.ndim == 1
for i in xrange(0, y.shape[0], cols):
print(y[i:i+cols])
pv.add_patch(display_batch[row_start+j,:,:,:], rescale = False)
if mapback:
pv.add_patch(mapped_batch[row_start+j,:,:,:], rescale = False)
pv.show()
if hasattr(model.visible_layer, 'beta'):
beta = model.visible_layer.beta.get_value()
#model.visible_layer.beta.set_value(beta * 100.)
print('beta: ',(beta.min(), beta.mean(), beta.max()))
print('showing seed data...')
show()
print('How many Gibbs steps should I run with the seed data clamped? (negative = ignore seed data) ')
x = int(input())
# Make shared variables representing the sampling state of the model
layer_to_state = model.make_layer_to_state(m)
# Seed the sampling with the data batch
vis_sample = layer_to_state[model.visible_layer]
def validate_all_samples():
# Run some checks on the samples, this should help catch any bugs
layers = [ model.visible_layer ] + model.hidden_layers
def check_batch_size(l):
if isinstance(l, (list, tuple)):
map(check_batch_size, l)
else:
rows = 5
cols = 10
m = rows * cols
_, model_path = sys.argv
print('Loading model...')
model = serial.load(model_path)
model.set_batch_size(m)
dataset_yaml_src = model.dataset_yaml_src
print('Loading data...')
dataset = yaml_parse.load(dataset_yaml_src)
x = input('use test set? (y/n) ')
if x == 'y':
dataset = dataset.get_test_set()
else:
assert x == 'n'
vis_batch = dataset.get_batch_topo(m)
_, patch_rows, patch_cols, channels = vis_batch.shape
assert _ == m
mapback = hasattr(dataset, 'mapback_for_viewer')
actual_cols = 2 * cols * (1 + mapback) * (1 + (channels == 2))
def remove_packages(packages_to_remove):
"""
Uninstall packages, whether or not they
are found in the source.lst (so it can
remove datasets installed from file).
:param packages_to_remove: list of package names
:raises: IOErrors
"""
if packages_to_remove == []:
raise RuntimeError("[rm] fatal: need packages names to remove.")
packages_really_to_remove = []
for this_package in packages_to_remove:
if this_package in packages_sources:
#this_data_set_location=os.path.join(dataset_data_path,this_package)
# check if in the installed.lst
# then if directory actually exists
# then if you have rights to remove it
if this_package in installed_packages_list:
this_data_set_location = os.path.join(
installed_packages_list[this_package].where,
installed_packages_list[this_package].name)
if os.path.exists(this_data_set_location):
if (file_access_rights(this_data_set_location, os.W_OK)):
# ok, you may have rights to delete it
packages_really_to_remove.append(this_package)
else:
logger.warning("[rm] insufficient rights "
"to remove '{0}'".format(this_package))
else:
logger.warning("[rm] package '{0}' found in config file "
"but not installed".format(this_package))
else:
logger.warning("[rm] package '{0}' "
"not installed".format(this_package))
else:
logger.warning("[rm] unknown package '{0}'".format(this_package))
if packages_really_to_remove != []:
logger.info("[rm] the following packages will be removed permanently:")
packages = []
for this_package in packages_really_to_remove:
packages.append(this_package)
logger.info(' '.join(packages))
r = input("Proceed? [yes/N] ")
if r == 'y' or r == 'yes':
for this_package in packages_really_to_remove:
remove_package(installed_packages_list[this_package],
dataset_data_path)
else:
logger.info("[up] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to remove, filenames where bad.
pass
def show_samples(m, model_path):
"""
Show samples given a DBM model.
Parameters
----------
m: int
rows * cols
model_path: str
Path of the model.
"""
model = load_model(model_path, m)
print('Loading data (used for setting up visualization '
'and seeding gibbs chain) ...')
dataset_yaml_src = model.dataset_yaml_src
dataset = yaml_parse.load(dataset_yaml_src)
pv = init_viewer(dataset, rows, cols)
if hasattr(model.visible_layer, 'beta'):
beta = model.visible_layer.beta.get_value()
print('beta: ', (beta.min(), beta.mean(), beta.max()))
print('showing seed data...')
vis_batch = dataset.get_batch_topo(m)
update_viewer(dataset, pv, vis_batch, rows, cols)
pv.show()
print('How many Gibbs steps should I run with the seed data clamped?'
'(negative = ignore seed data)')
x = int(input())
# Make shared variables representing the sampling state of the model
layer_to_state = model.make_layer_to_state(m)
# Seed the sampling with the data batch
vis_sample = layer_to_state[model.visible_layer]
validate_all_samples(model, layer_to_state)
if x >= 0:
if vis_sample.ndim == 4:
vis_sample.set_value(vis_batch)
else:
design_matrix = dataset.get_design_matrix(vis_batch)
vis_sample.set_value(design_matrix)
validate_all_samples(model, layer_to_state)
sample_func = get_sample_func(model, layer_to_state, x)
while True:
print('Displaying samples. '
'How many steps to take next? (q to quit, ENTER=1)')
while True:
x = input()
if x == 'q':
quit()
if x == '':
x = 1
break
else:
try:
x = int(x)
break
except ValueError:
print('Invalid input, try again')
for i in xrange(x):
print(i)
sample_func()
validate_all_samples(model, layer_to_state)
vis_batch = vis_sample.get_value()
update_viewer(dataset, pv, vis_batch, rows, cols)
pv.show()
if 'Softmax' in str(type(model.hidden_layers[-1])):
state = layer_to_state[model.hidden_layers[-1]]
value = state.get_value()
y = np.argmax(value, axis=1)
assert y.ndim == 1
for i in xrange(0, y.shape[0], cols):
print(y[i:i + cols])
def upgrade_packages(packages_to_upgrade, hook=None):
"""
Upgrades packages.
If no packages are supplied, it will perform
an "update-all" operation, finding all packages
that are out of date.
If packages names are supplied, only those
are checked for upgrade (and upgraded if out
of date)
:param packages_to_upgrade: list of package names.
:raises: IOErrors (from downloads/rights)
"""
# get names only
if packages_to_upgrade == []:
packages_to_upgrade = installed_packages_list.keys() # all installed!
all_packages = True
else:
all_packages = False
# check what packages are in the list,
# and really to be upgraded.
#
packages_really_to_upgrade = []
for this_package in packages_to_upgrade:
if this_package in installed_packages_list:
# check if there's a date
installed_date = installed_packages_list[this_package].timestamp
if this_package in packages_sources:
repo_date = packages_sources[this_package].timestamp
if installed_date < repo_date:
# ok, there's a newer version
logger.info(this_package)
packages_really_to_upgrade.append(this_package)
else:
# no newer version, nothing to update
pass
else:
logger.warning("[up] '{0}' is unknown "
"(installed from file?).".format(this_package))
else:
# not installed?
if not all_packages:
logger.warning("[up] '{0}' is not installed, "
"cannot upgrade.".format(this_package))
pass
# once we have determined which packages
# are to be updated, we show them to the
# user for him to confirm
#
if packages_really_to_upgrade != []:
logger.info("[up] the following package(s) will be upgraded:")
for this_package in packages_really_to_upgrade:
readable_size = packages_sources[this_package].readable_size
logger.info("{0} ({1})".format(this_package, readable_size))
r = input("Proceed? [yes/N] ")
if r == 'y' or r == 'yes':
for this_package in packages_really_to_upgrade:
install_upgrade(packages_sources[this_package],
upgrade=True,
progress_hook=hook)
else:
logger.info("[up] Taking '{0}' for no, so there.".format(r))
else:
# ok, nothing to upgrade,
# move along.
pass
def get_weights_report(model_path=None,
model=None,
rescale='individual',
border=False,
norm_sort=False,
dataset=None):
"""
Returns a PatchViewer displaying a grid of filter weights
Parameters
----------
model_path : str
Filepath of the model to make the report on.
rescale : str
A string specifying how to rescale the filter images: \
'individual' (default): scale each filter so that it \
uses as much as possible of the dynamic range \
of the display under the constraint that 0 \
is gray and no value gets clipped \
'global' : scale the whole ensemble of weights \
'none' : don't rescale
dataset: pylearn2.datasets.dataset.Dataset
Dataset object to do view conversion for displaying the weights. If \
not provided one will be loaded from the model's dataset_yaml_src.
Returns
-------
WRITEME
"""
if model is None:
print 'making weights report'
print 'loading model'
model = serial.load(model_path)
print 'loading done'
else:
assert model_path is None
assert model is not None
if rescale == 'none':
global_rescale = False
patch_rescale = False
elif rescale == 'global':
global_rescale = True
patch_rescale = False
elif rescale == 'individual':
global_rescale = False
patch_rescale = True
else:
raise ValueError('rescale=' + rescale +
", must be 'none', 'global', or 'individual'")
if hasattr(model, 'layers'):
if isinstance(model.layers[0], mlp_models.PretrainedLayer):
model = model.layers[0].layer_content
if isinstance(model, dict):
#assume this was a saved matlab dictionary
del model['__version__']
del model['__header__']
del model['__globals__']
keys = [key for key in model \
if hasattr(model[key], 'ndim') and model[key].ndim == 2]
if len(keys) > 2:
key = None
while key not in keys:
logger.info('Which is the weights?')
for key in keys:
logger.info('\t{0}'.format(key))
key = input()
else:
key, = keys
weights = model[key]
norms = np.sqrt(np.square(weights).sum(axis=1))
print 'min norm: ',norms.min()
print 'mean norm: ',norms.mean()
print 'max norm: ',norms.max()
return patch_viewer.make_viewer(weights,
is_color=weights.shape[1] % 3 == 0)
weights_view = None
W = None
try:
weights_view = model.get_weights_topo()
h = weights_view.shape[0]
except NotImplementedError:
if dataset is None:
print 'loading dataset...'
control.push_load_data(False)
dataset = yaml_parse.load(model.dataset_yaml_src)
control.pop_load_data()
print '...done'
try:
W = model.get_weights()
#print 'recons: '
#for ch in xrange(3):
# chv = r[:,:,ch]
# print '\t',ch,(chv.min(),chv.mean(),chv.max())
if mapback:
pv.add_patch(dataset.adjust_to_be_viewed_with(
mapped_r_batch[row_start+j,:,:,:].copy(),
mapped_batch[row_start+j,:,:,:].copy()),rescale = False)
pv.show()
if hasattr(grbm.visible_layer, 'beta'):
beta = grbm.visible_layer.beta.get_value()
#model.visible_layer.beta.set_value(beta * 100.)
print('beta: ',(beta.min(), beta.mean(), beta.max()))
while True:
show()
print('Displaying reconstructions. (q to quit, ENTER = show more)')
while True:
x = input()
if x == 'q':
quit()
if x == '':
x = 1
break
else:
print('Invalid input, try again')
vis_batch = dataset.get_batch_topo(m)
