def get_block(self):
self.source = GenSource(e
for i in xrange(len(self.cross_val_keys))
if not i == self.cross_val_index
for e in self.data[self.cross_val_keys[i]])
self.ph_source = Transformer([self.source], {self.source.getName(): 'd'}, lambda d: d['phase'])
self.bw_source = Transformer([self.source], {self.source.getName(): 'd'}, lambda d: d['raw'])
block = SignalBlock(
[self.source],
[self.bw_source, self.ph_source]
)
self.source.callback = block.stop
return block
class NormSourceMaker(object):
""" An experiment, not really useful, keeping the code around for basically no reason """
def __init__(self, datafolder = None, phases = None, cross_val = False):
self.data = None
if datafolder:
with gzip.open(datafolder+"normalized.pkl.gz",'rb') as f:
self.data = cPickle.load(f)
else:
paths = glob.glob("~/PycharmProjects/MatchBrain/ml/normalized.pkl.gz")
if len(paths) > 0:
with gzip.open(paths[0]) as f:
self.data = cPickle.load(f)
if not self.data:
return
self.phases = phases or phase_names
self.cross_val_keys = list(self.data.keys())
self.cross_val_index = 0 if cross_val else None
def cross_val_next(self):
self.cross_val_index = (self.cross_val_index + 1) % len(self.cross_val_keys)
def get_block(self):
self.source = GenSource(e
for i in xrange(len(self.cross_val_keys))
if not i == self.cross_val_index
for e in self.data[self.cross_val_keys[i]])
self.ph_source = Transformer([self.source], {self.source.getName(): 'd'}, lambda d: d['phase'])
self.bw_source = Transformer([self.source], {self.source.getName(): 'd'}, lambda d: d['raw'])
block = SignalBlock(
[self.source],
[self.bw_source, self.ph_source]
)
self.source.callback = block.stop
return block
def get_block(self, shift = 1):
a_meas = self.raws_per_phase[0][1]
self.ph_source = GenSource(li[0]
for rep in repeat(self.raws_per_phase, len(a_meas))
for li in rep)
self.bw_source = GenSource(li
for i in xrange(0, len(a_meas), shift)
for li in map(lambda t: interp_ls(t[0][1], t[1][1], i)
,zip(self.raws_per_phase[0:-1], self.raws_per_phase[1:])))
block = SignalBlock(
[self.bw_source, self.ph_source],
[Transformer(
[self.bw_source, self.ph_source],
{self.bw_source.getName(): 'b', self.ph_source.getName(): 'd'},
lambda b,d: (b,d))
]
)
self.bw_source.callback = block.stop
self.ph_source.callback = block.stop
#result.sinks.append(Sink([self.bw_source, self.ph_source], lambda x: print(x)))
return block
class LogSourceMaker(object):
""" The way the system used to work, this hooks into the whole signal system thing I built for the game code, not useful for learning """
def __init__(self, clean_seconds = 3, logfolder = None, phases = None, cross_val = False):
if logfolder:
all_dict = loadall(logfolder=logfolder)
else:
all_dict = loadall()
self.seen_val_names = []
if cross_val:
def tname_for(fname):
just_f = fname.split('/')[-1]
return just_f.split('2015')[0]
#tester_names = sorted(list(set(map(tname_for, all_dict.iterkeys()))))
#print(tester_names)
#print({f: max(g) for f,g in groupby(sorted(all_dict.iterkeys()), tname_for)})
all_dict = {f: all_dict[max(g)]
for f,g in groupby(sorted(all_dict.iterkeys()), tname_for)}
val_name = random.choice(all_dict.keys())
self.seen_val_names.append(val_name)
self.val_dict = {val_name: all_dict[val_name]}
self.all_dict = {k:v for (k,v) in all_dict.iteritems() if not k is val_name}
else:
self.val_dict = {}
self.all_dict = all_dict
self.phases = phases or phase_names
#self.phases.append('none')
self.clean_seconds = clean_seconds
self.raws_per_phase = self.to_raws_per_phase(all_dict)
self.rpp_val = self.to_raws_per_phase(self.val_dict)
def cross_val_next(self):
all_dict = self.all_dict
all_dict.update(self.val_dict)
next_val = random.choice(all_dict.keys())
while next_val in self.seen_val_names:
next_val = random.choice(all_dict.keys())
self.val_dict = {next_val: all_dict[next_val]}
self.all_dict = {k:v for (k,v) in all_dict.iteritems() if not k is next_val}
self.raws_per_phase = self.to_raws_per_phase(self.all_dict)
self.rpp_val = self.to_raws_per_phase(self.val_dict)
def to_raws_per_phase(self, a_dict):
result = []
cur_phase = phase_names[0]
for p in a_dict.keys():
cur_part = a_dict[p]
seen_phase = False
clean_counter = self.clean_seconds
for li in cur_part:
if "data" in li:
if "train" in li["data"] and li["data"]["train"]["phase"] in self.phases:
cur_phase = li["data"]["train"]["phase"]
seen_phase = True
clean_counter = self.clean_seconds
elif "brainwave" in li["data"]:
if clean_counter > 0:
clean_counter -= 1
else:
result.append((cur_phase
if seen_phase
else "DISTRACT"
,li["data"]["brainwave"]["raw"]))
else:
# Skip data for "phases" that aren't in the list
pass
return result
def get_block(self, shift = 1):
a_meas = self.raws_per_phase[0][1]
self.ph_source = GenSource(li[0]
for rep in repeat(self.raws_per_phase, len(a_meas))
for li in rep)
self.bw_source = GenSource(li
for i in xrange(0, len(a_meas), shift)
for li in map(lambda t: interp_ls(t[0][1], t[1][1], i)
,zip(self.raws_per_phase[0:-1], self.raws_per_phase[1:])))
block = SignalBlock(
[self.bw_source, self.ph_source],
[Transformer(
[self.bw_source, self.ph_source],
{self.bw_source.getName(): 'b', self.ph_source.getName(): 'd'},
lambda b,d: (b,d))
]
)
self.bw_source.callback = block.stop
self.ph_source.callback = block.stop
#result.sinks.append(Sink([self.bw_source, self.ph_source], lambda x: print(x)))
return block
