Predict circular estimate statistics

author: steeve laquitaine

This tutorial simulates the statistics (mean and standard deviations) of the circular estimate predictions generated by a cardinal Bayesian model.

Setup

# go to the project's root path
import os
os.chdir("..")

# import dependencies
from bsfit.nodes.models.bayes import CardinalBayes
from bsfit.nodes.dataEng import (
    simulate_dataset,
)
from matplotlib import pyplot as plt
%load_ext autoreload
%autoreload 2

The autoreload extension is already loaded. To reload it, use:
  %reload_ext autoreload

Set the parameters

# set the parameters
PRIOR_SHAPE = "vonMisesPrior"
PRIOR_MODE = 225
OBJ_FUN = "maxLLH"
READOUT = "map"
PRIOR_NOISE = [80, 40]      # e.g., prior's std
STIM_NOISE = [0.33, 0.66, 1.0]
SIM_P = {
    "k_llh": [2.7, 10.7, 33],
    "k_prior": [2.7, 33],
    "k_card": [2000],
    "prior_tail": [0],
    "p_rand": [0],
    "k_m": [2000],
}
GRANULARITY = "mean"
CENTERING = True

Simulate a dataset

# simulate a training dataset
train_dataset = simulate_dataset(
    stim_noise=STIM_NOISE,
    prior_mode=PRIOR_MODE,
    prior_noise=PRIOR_NOISE,
    prior_shape=PRIOR_SHAPE,
)

Simulate the model mean predictions

# instantiate the model
model = CardinalBayes(
    initial_params=SIM_P,
    prior_shape=PRIOR_SHAPE,
    prior_mode=PRIOR_MODE,
    readout=READOUT
    )

# simulate predictions
plt.figure(figsize=(15,5))
model = model.simulate(
    dataset=train_dataset,
    granularity=GRANULARITY,
    centering=CENTERING,
)

Running simulation ...

Calculating predictions ...

-logl:2537.18, aic:5092.36, kl:[ 2.7 10.7 33. ], kp:[ 2.7 33. ], kc:[2000.], pt:0.00, pr:0.00, km:2000.00

Tutorial complete !