def eval_conditional_probability(x, parameters, exercise_ind):
"""Evaluate the conditional probability of answering each question
accurately for a student with ability x
"""
return mirt_util.conditional_probability_correct(
np.ones((parameters.num_abilities, 1)) * x, parameters,
exercise_ind)
def eval_conditional_probability(x, parameters, exercise_ind):
"""Evaluate the conditional probability of answering each question
accurately for a student with ability x
"""
return mirt_util.conditional_probability_correct(
np.ones((parameters.num_abilities, 1)) * x,
parameters,
exercise_ind)
def estimated_exercise_accuracy(self, history, exercise_name,
update_abilities=True, ignore_analytics=False):
"""Returns the expected probability of getting a future question
correct on the specified exercise.
"""
if update_abilities:
self._update_abilities(history, ignore_analytics=ignore_analytics)
try:
exercise_ind = mirt_util.get_exercise_ind(
exercise_name, self.exercise_ind_dict)
except KeyError:
# If we don't have this exercise, predict the mean predicted
# accuracy over all exercises we do have.
return self.score(history)
return mirt_util.conditional_probability_correct(
self.abilities, self.theta, exercise_ind)[0]
def estimated_exercise_accuracy(self,
history,
exercise_name,
update_abilities=True,
ignore_analytics=False):
"""Returns the expected probability of getting a future question
correct on the specified exercise.
"""
if update_abilities:
self._update_abilities(history, ignore_analytics=ignore_analytics)
try:
exercise_ind = mirt_util.get_exercise_ind(exercise_name,
self.exercise_ind_dict)
except KeyError:
# If we don't have this exercise, predict the mean predicted
# accuracy over all exercises we do have.
return self.score(history)
return mirt_util.conditional_probability_correct(
self.abilities, self.theta, exercise_ind)[0]
