def try_to_update_cur_outcome(self):
print("Entering try_to_update_cur_outcome...")
print(" current effect: %s" % self.cur_effect)
e1, n1, e2, n2 = self.ma_unit.get_raw_data_for_groups(self.cur_groups)
print(" e1: %s, n1: %s, e2: %s, n2: %s" % (str(e1),str(n1),str(e2),str(n2)))
two_arm_raw_data_ok = not any([self._isBlank(x) for x in [e1, n1, e2, n2]])
one_arm_raw_data_ok = not any([self._isBlank(x) for x in [e1, n1]])
curr_effect_is_one_arm = self.cur_effect in BINARY_ONE_ARM_METRICS
curr_effect_is_two_arm = self.cur_effect in BINARY_TWO_ARM_METRICS
# if None is in the raw data, should we clear out current outcome?
if two_arm_raw_data_ok or (curr_effect_is_one_arm and one_arm_raw_data_ok):
if curr_effect_is_two_arm:
est_and_ci_d = meta_py_r.effect_for_study(e1, n1, e2, n2, metric=self.cur_effect, conf_level=self.CI_spinbox.value())
else:
# binary, one-arm
est_and_ci_d = meta_py_r.effect_for_study(e1, n1, two_arm=False, metric=self.cur_effect, conf_level=self.CI_spinbox.value())
display_est, display_low, display_high = est_and_ci_d["display_scale"]
self.ma_unit.set_display_effect_and_ci(self.cur_effect, self.group_str, display_est, display_low, display_high)
est, low, high = est_and_ci_d["calc_scale"] # calculation (e.g., log) scale
self.ma_unit.set_effect_and_ci(self.cur_effect, self.group_str, est, low, high)
self.set_current_effect()
def try_to_update_cur_outcome(self):
e1, n1, e2, n2 = self.ma_unit.get_raw_data_for_groups(self.cur_groups)
print("e1: %s, n1: %s, e2: %s, n2: %s" %
(str(e1), str(n1), str(e2), str(n2)))
two_arm_raw_data_ok = not any(
[self._isBlank(x) for x in [e1, n1, e2, n2]])
one_arm_raw_data_ok = not any([self._isBlank(x) for x in [e1, n1]])
curr_effect_is_one_arm = self.cur_effect in BINARY_ONE_ARM_METRICS
curr_effect_is_two_arm = self.cur_effect in BINARY_TWO_ARM_METRICS
# if None is in the raw data, should we clear out current outcome?
if two_arm_raw_data_ok or (curr_effect_is_one_arm
and one_arm_raw_data_ok):
if curr_effect_is_two_arm:
est_and_ci_d = meta_py_r.effect_for_study(
e1,
n1,
e2,
n2,
metric=self.cur_effect,
conf_level=self.global_conf_level)
else:
# binary, one-arm
est_and_ci_d = meta_py_r.effect_for_study(
e1,
n1,
two_arm=False,
metric=self.cur_effect,
conf_level=self.global_conf_level)
est, low, high = est_and_ci_d[
"calc_scale"] # calculation (e.g., log) scale
self.ma_unit.set_effect_and_ci(self.cur_effect,
self.group_str,
est,
low,
high,
mult=self.mult)
self.set_current_effect()
def update_outcome_if_possible(self, study_index):
"""
Checks the parametric study to ascertain if enough raw data has been
entered to compute the outcome. If so, the outcome is computed and
displayed.
"""
est, lower, upper = None, None, None
data_type = self.get_current_outcome_type(get_str=False)
if self.raw_data_is_complete_for_study(study_index):
if data_type == BINARY:
e1, n1, e2, n2 = self.get_cur_raw_data_for_study(study_index)
est, lower, upper = meta_py_r.effect_for_study(e1, n1, e2, n2)
elif data_type == CONTINUOUS:
n1, m1, se1, n2, m2, se2 = self.get_cur_raw_data_for_study(study_index)
est, lower, upper = meta_py_r.continuous_effect_for_study(n1, m1, se1, n2, m2, se2)
ma_unit = self.get_current_ma_unit_for_study(study_index)
# now set the effect size & CIs
ma_unit.set_effect_and_ci(self.current_effect, est, lower, upper)