def feal_atemp(cls, rms, scaleto=10):
"""Atemporal (individual frame) featue landscape
"""
log_reld = op.makeLogisticFunc(scaleto, -3, 0)
# Counts (feature 0..4)
fealand = [0 for i in range(5)]
fealand[np.argmin(rms)] = scaleto
tup = []
# Proximity (feature 5..9)
# Normalized and adjusted to smooth implicit water
A, B = np.argsort(rms)[:2]
prox = op.makeLogisticFunc(scaleto, scaleto, (rms[B]+rms[A])/2)
for d in rms:
tup.append(prox(d))
# Relative Distance (akin to LLE) (feature 10..19)
for a in range(4):
for b in range(a+1, 5):
rel_dist = rms[a]-rms[b]
tup.append(log_reld(rel_dist))
fealand.extend(tup)
# Additional Feature Spaces Would go here
return np.array(fealand) # Tuple or NDArray?
def feal_atemp(rms, scaleto=10):
"""Atemporal (individual frame) featue landscape
"""
log_reld = op.makeLogisticFunc(scaleto, -3, 0)
# Counts (feature 0..4)
fealand = [0 for i in range(5)]
fealand[np.argmin(rms)] = scaleto
tup = []
# Proximity (feature 5..9)
for n, dist in enumerate(rms):
# tup.append(log_prox(dist))
maxd = 10. #11.34
# tup.append(scaleto*max(maxd-dist, 0)/maxd)
tup.append(max(maxd-dist, 0))
# Relative Distance (akin to LLE) (feature 10..19)
for a in range(4):
for b in range(a+1, 5):
rel_dist = rms[a]-rms[b]
tup.append(log_reld(rel_dist))
fealand.extend(tup)
# Additional Feature Spaces Would go here
return np.array(fealand) # Tuple or NDArray?
def atemporal2(cls, rms, scaleto=10):
"""Atemporal (individual frame) featue landscape
"""
log_reld = op.makeLogisticFunc(scaleto, -3, 0)
maxd = 20
mind = 10
fealand = [0 for i in range(5)]
fealand[np.argmin(rms)] = scaleto
tup = []
# Proximity
for dist in rms:
fealand.append(scaleto * max(maxd - (max(dist, mind)), 0) /
(maxd - mind))
# Additional Feature Spaces
for a in range(4):
for b in range(a + 1, 5):
rel_dist = rms[a] - rms[b]
tup.append(log_reld(rel_dist))
fealand.extend(tup)
# Additional Feature Spaces Would go here
return np.array(fealand) # Tuple or NDArray?
def atemporal2(cls, rms, scaleto=10):
"""Atemporal (individual frame) featue landscape
"""
log_reld = op.makeLogisticFunc(scaleto, -3, 0)
maxd = 20
mind = 10
fealand = [0 for i in range(5)]
fealand[np.argmin(rms)] = scaleto
tup = []
# Proximity
for dist in rms:
fealand.append(scaleto*max(maxd-(max(dist, mind)), 0)/(maxd-mind))
# Additional Feature Spaces
for a in range(4):
for b in range(a+1, 5):
rel_dist = rms[a]-rms[b]
tup.append(log_reld(rel_dist))
fealand.extend(tup)
# Additional Feature Spaces Would go here
return np.array(fealand) # Tuple or NDArray?
basin_score_well = np.zeros(91116)
basin_score_tran = np.zeros(91116)
multi_basins = []
max_sigma = np.max(sigma)
cluster_score_well, cluster_score_tran = np.zeros(len(cluk)), np.zeros(
len(cluk))
C_ws, C_wv = .4, .6
C_ts, C_tv = .1, .90
B_wd, B_wv = .9, .1
B_td, B_tv = .25, .75
max_var1 = np.max(variance[np.nonzero(variance)])
max_var2 = np.max(variance2[np.nonzero(variance2)])
sizeFunc = op.makeLogisticFunc(2, 0.001, 91116 / len(clulist))
logging.info("CLUSTER_SUMMARY %d %d", support, num_clu)
for n, k in enumerate(cluk):
cSize = len(clulist[n])
cVar = variance2[n]
sc_var = 0 if cVar == 0 else cVar / max_var2
sc_size = sizeFunc(
cSize) - 1 # max(-2, 1 - (len(cluk)*cSize / (91116)))
cluster_score_well[n] = max(0, C_wv * (1 - sc_var) + C_ws * sc_size)
cluster_score_tran[n] = max(0, C_tv * sc_var + C_ts * sc_size)
# cnt = np.bincount(list(it.chain(*[[c for c in dL[i]] for i in clulist[n]])))
cnt = np.bincount([getstate(i) for i in clulist[n]], minlength=6)
logging.info(
'%2d %15s | sz=%6d | var=%7.2f | scW=%5.2f scT%5.2f) %s' %
(n, k, cSize, cVar, cluster_score_well[n], cluster_score_tran[n],
cnt / sum(cnt)))
basin_score_well = np.zeros(91116)
basin_score_tran = np.zeros(91116)
multi_basins = []
max_sigma = np.max(sigma)
cluster_score_well, cluster_score_tran = np.zeros(len(cluk)), np.zeros(len(cluk))
C_ws, C_wv = .4, .6
C_ts, C_tv = .1, .90
B_wd, B_wv = .9, .1
B_td, B_tv = .25, .75
max_var1 = np.max(variance[np.nonzero(variance)])
max_var2 = np.max(variance2[np.nonzero(variance2)])
sizeFunc = op.makeLogisticFunc(2, 0.001, 91116/len(clulist))
logging.info("CLUSTER_SUMMARY %d %d", support, num_clu)
for n, k in enumerate(cluk):
cSize = len(clulist[n])
cVar = variance2[n]
sc_var = 0 if cVar == 0 else cVar / max_var2
sc_size = sizeFunc(cSize) -1 # max(-2, 1 - (len(cluk)*cSize / (91116)))
cluster_score_well[n] = max(0, C_wv *(1-sc_var) + C_ws * sc_size)
cluster_score_tran[n] = max(0, C_tv * sc_var + C_ts * sc_size)
# cnt = np.bincount(list(it.chain(*[[c for c in dL[i]] for i in clulist[n]])))
cnt = np.bincount([getstate(i) for i in clulist[n]], minlength=6)
logging.info('%2d %15s | sz=%6d | var=%7.2f | scW=%5.2f scT%5.2f) %s' %
(n, k, cSize, cVar, cluster_score_well[n], cluster_score_tran[n], cnt/sum(cnt)))
max_dist = np.max([s for _,s in elmlist[n]])
max_sigma = np.max([sigma[i] for i,_ in elmlist[n]])
for i, s in elmlist[n]:
