def count_ps_from_beta(N, beta, verbose=False, A=4, log_cols=None, entropies=None):
iterator = (lambda x:tqdm(x, total=partitions(N, A))) if verbose else (lambda x:x)
if log_cols is None or entropies is None:
log_ws = np.array([log_counts_to_cols(count, A=A) + (-beta*entropy_from_counts(count))
for count in iterator(enumerate_counts_iter(N, A=A))])
else:
log_ws = log_cols + -beta*entropies
return np.exp(np_log_normalize(log_ws))
def count_ps_from_beta(N,
beta,
verbose=False,
A=4,
log_cols=None,
entropies=None):
iterator = (lambda x: tqdm(x, total=partitions(N, A))) if verbose else (
lambda x: x)
if log_cols is None or entropies is None:
log_ws = np.array([
log_counts_to_cols(count, A=A) +
(-beta * entropy_from_counts(count))
for count in iterator(enumerate_counts_iter(N, A=A))
])
else:
log_ws = log_cols + -beta * entropies
return np.exp(np_log_normalize(log_ws))
def f2(beta):
log_phats = np_log_normalize(log_cols + -beta * entropies)
expected_entropy = np.exp(log_phats).dot(entropies)
return log2(A) - expected_entropy - desired_ic_per_col
def f2(beta):
log_phats = np_log_normalize(log_cols + -beta*entropies)
expected_entropy = np.exp(log_phats).dot(entropies)
return log2(A) - expected_entropy - desired_ic_per_col