def loglike(self, beliefs, meta, state, var_value, params, latents, pos):
assert self.name in ['own_noise', 'meta_noise']
beliefs_indiv = [bel[:, pos:pos+1] for bel in beliefs]
meta_indiv = [met[:, pos:pos+1] for met in meta]
var_value = [var_value]
expertise = get_expertise(state, params.use_expertise)
if self.name == "own_noise":
rll = [scoring.reports_ll_marg(bel, met, lat['world'],
lat['sm'], lat['wp'],
var_value,
state.meta_noise[pos:pos+1],
expertise,
params, lat['bm'],
lat['mm'])
for bel, met, lat in
zip(beliefs_indiv, meta_indiv,
latents)]
elif self.name == "meta_noise":
rll = [scoring.reports_ll_marg(bel, met, lat['world'],
lat['sm'], lat['wp'],
state.own_noise[pos:pos+1],
var_value,
expertise,
params, lat['bm'],
lat['mm'])
for bel, met, lat in
zip(beliefs_indiv, meta_indiv,
latents)]
return np.sum(rll)
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
assert self.name in ['own_noise', 'meta_noise']
expertise = get_expertise(latents, params.use_expertise)
if self.name == "own_noise":
rll = scoring.reports_ll_marg(beliefs, meta, state.world,
state.sm, state.wp,
var_value, state.meta_noise,
expertise, params)
elif self.name == "meta_noise":
rll = scoring.reports_ll_marg(beliefs, meta, state.world,
state.sm, state.wp,
state.own_noise, var_value,
expertise, params)
return rll
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
lpw = scoring.log_prob_world(state.wp, var_value)
own_noise, meta_noise = get_noise(state, latents, self.noise_in_state)
expertise = get_expertise(latents, params.use_expertise)
rll = scoring.reports_ll_marg(beliefs, meta, var_value, state.sm,
state.wp, own_noise, meta_noise,
expertise, params)
return lpw + rll
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
#assert var_value[0, 0] > var_value[0, 1], "Invalid signal matrix"
lp_sm = 0
own_noise, meta_noise = get_noise(state, latents, self.noise_in_state)
expertise = get_expertise(latents, params.use_expertise)
rll = scoring.reports_ll_marg(beliefs, meta, state.world, var_value,
state.wp, own_noise, meta_noise,
expertise, params)
return lp_sm + rll
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
if var_value[0] > 1 or var_value[0] < 0:
return -np.inf
#Original code had checks for bijection and kind of prior.
lpw = scoring.log_prob_world(var_value, state.world)
expertise = get_expertise(latents, params.use_expertise)
own_noise, meta_noise = get_noise(state, latents, self.noise_in_state)
rll = scoring.reports_ll_marg(beliefs, meta, state.world, state.sm,
var_value, own_noise, meta_noise,
expertise, params)
return lpw + rll
def loglike(self, beliefs, meta, state, var_value, params, latents, pos):
beliefs_indiv = [bel[:, pos:pos+1] for bel in beliefs]
meta_indiv = [met[:, pos:pos+1] for met in meta]
var_value = [var_value]
rll = [scoring.reports_ll_marg(bel, met, lat['world'], lat['sm'],
lat['wp'],
state['own_noise'][pos:pos+1],
state['meta_noise'][pos:pos+1],
var_value, params, lat['bm'], lat['mm'])
for bel, met, lat in
zip(beliefs_indiv, meta_indiv, latents)]
return np.sum(rll)
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
assert self.name in ["own_noise", "meta_noise"]
expertise = get_expertise(latents, params.use_expertise)
bias = get_bias(latents, params.use_bias)
lp = self.noise_log_prior(var_value)
if self.name == "own_noise":
own_noise = var_value
meta_noise = state.meta_noise
else:
own_noise = state.own_noise
meta_noise = var_value
rll = scoring.reports_ll_marg(
beliefs, meta, state.world, state.sm, state.wp, own_noise, meta_noise, expertise, bias, params
)
return lp + rll
def loglike(self, beliefs, meta, state, var_value, params, latents, _):
# assert var_value[0, 0] > var_value[0, 1], "Invalid signal matrix"
lp_sm = 0
own_noise, meta_noise = get_noise(state, latents, self.noise_in_state)
expertise = get_expertise(latents, params.use_expertise)
bias = get_bias(latents, params.use_bias)
# sm = np.array([[.66, .34], [.34, .66]])
# wp = np.array([.43, .57])
# world = 1
# sm0 = np.array([[.7, .36], [.3, .64]])
# wp0 = np.array([.56, .44])
# world0 = 0
# blah = scoring.reports_ll_person(beliefs, meta, world, sm, wp, own_noise,
# meta_noise, expertise, bias, params, debug=1)
rll = scoring.reports_ll_marg(
beliefs, meta, state.world, var_value, state.wp, own_noise, meta_noise, expertise, bias, params
)
return lp_sm + rll
