Python numpy.nanargmax方法代码示例

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def _split(
    data: pd.DataFrame, value: Union[str, int], unit: Optional[str]
) -> Iterator[pd.DataFrame]:
    # This method helps splitting a flight into several.
    if data.shape[0] < 2:
        return
    diff = data.timestamp.diff().values
    if unit is None:
        delta = pd.Timedelta(value).to_timedelta64()
    else:
        delta = np.timedelta64(value, unit)
    # There seems to be a change with numpy >= 1.18
    # max() now may return NaN, therefore the following fix
    max_ = np.nanmax(diff)
    if max_ > delta:
        # np.nanargmax seems bugged with timestamps
        argmax = np.where(diff == max_)[0][0]
        yield from _split(data.iloc[:argmax], value, unit)
        yield from _split(data.iloc[argmax:], value, unit)  # noqa
    else:
        yield data


# flake B008 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def calculate_cis_permutations(genotypes_t, phenotype_t, permutation_ix_t, residualizer=None):
    """Calculate nominal and empirical correlations"""
    permutations_t = phenotype_t[permutation_ix_t]

    r_nominal_t, genotype_var_t, phenotype_var_t = calculate_corr(genotypes_t, phenotype_t.reshape(1,-1),
                                                                  residualizer=residualizer, return_var=True)
    std_ratio_t = torch.sqrt(phenotype_var_t.reshape(1,-1) / genotype_var_t.reshape(-1,1))
    r_nominal_t = r_nominal_t.squeeze(dim=-1)
    std_ratio_t = std_ratio_t.squeeze(dim=-1)
    corr_t = calculate_corr(genotypes_t, permutations_t, residualizer=residualizer).pow(2)  # genotypes x permutations
    corr_t = corr_t[~torch.isnan(corr_t).any(1),:]
    if corr_t.shape[0] == 0:
        raise ValueError('All correlations resulted in NaN. Please check phenotype values.')
    r2_perm_t,_ = corr_t.max(0)  # maximum correlation across permutations

    r2_nominal_t = r_nominal_t.pow(2)
    r2_nominal_t[torch.isnan(r2_nominal_t)] = -1  # workaround for nanargmax()
    ix = r2_nominal_t.argmax()
    return r_nominal_t[ix], std_ratio_t[ix], ix, r2_perm_t, genotypes_t[ix] 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def save_everything(args, metrics_hist_all, model, model_dir, params, criterion, evaluate=False):
    """
        Save metrics, model, params all in model_dir
    """
    save_metrics(metrics_hist_all, model_dir)
    params['model_dir'] = model_dir
    save_params_dict(params)

    if not evaluate:
        #save the model with the best criterion metric
        if not np.all(np.isnan(metrics_hist_all[0][criterion])):
            if criterion == 'loss_dev': 
                eval_val = np.nanargmin(metrics_hist_all[0][criterion])
            else:
                eval_val = np.nanargmax(metrics_hist_all[0][criterion])

            if eval_val == len(metrics_hist_all[0][criterion]) - 1:                

		#save state dict
                sd = model.cpu().state_dict()
                torch.save(sd, model_dir + "/model_best_%s.pth" % criterion)
                if args.gpu:
                    model.cuda()
    print("saved metrics, params, model to directory %s\n" % (model_dir)) 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def __get_next_center(self, centers):
        """!
        @brief Calculates the next center for the data.

        @param[in] centers (array_like): Current initialized centers represented by indexes.

        @return (array_like) Next initialized center.<br>
                (uint) Index of next initialized center if return_index is True.

        """

        distances = self.__calculate_shortest_distances(self.__data, centers)

        if self.__candidates == kmeans_plusplus_initializer.FARTHEST_CENTER_CANDIDATE:
            for index_point in centers:
                distances[index_point] = numpy.nan
            center_index = numpy.nanargmax(distances)
        else:
            probabilities = self.__calculate_probabilities(distances)
            center_index = self.__get_probable_center(distances, probabilities)

        return center_index 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def early_stop_decision(self, epoch, val_metric, val_loss):
        '''
	Stop training if validation loss has stopped decreasing and
	validation BLEU score has not increased for --patience epochs.

        WARNING: quits with sys.exit(0).

	TODO: this doesn't yet support early stopping based on TER
        '''

	if val_loss < self.best_val_loss:
	    self.wait = 0
        elif val_metric > self.best_val_metric or self.args.no_early_stopping:
            self.wait = 0
        else:
            self.wait += 1
            if self.wait >= self.patience:
                # we have exceeded patience
                if val_loss > self.best_val_loss:
                    # and loss is no longer decreasing
                    logger.info("Epoch %d: early stopping", epoch)
                    handle = open("checkpoints/%s/summary"
                                  % self.args.run_string, "a")
                    handle.write("Early stopping because patience exceeded\n")
                    best_bleu = np.nanargmax(self.val_metric)
                    best_loss = np.nanargmin(self.val_loss)
                    logger.info("Best Metric: %d | val loss %.5f score %.2f",
                                best_bleu+1, self.val_loss[best_bleu],
                                self.val_metric[best_bleu])
                    logger.info("Best loss: %d | val loss %.5f score %.2f",
                                best_loss+1, self.val_loss[best_loss],
                                self.val_metric[best_loss])
                    handle.close()
                    sys.exit(0) 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def log_performance(self):
        '''
        Record model performance so far, based on validation loss.
        '''
        handle = open("checkpoints/%s/summary" % self.args.run_string, "w")

        for epoch in range(len(self.val_loss)):
            handle.write("Checkpoint %d | val loss: %.5f bleu %.2f\n"
                         % (epoch+1, self.val_loss[epoch],
                            self.val_metric[epoch]))

        logger.info("---")  # break up the presentation for clarity

        # BLEU is the quickest indicator of performance for our task
        # but loss is our objective function
        best_bleu = np.nanargmax(self.val_metric)
        best_loss = np.nanargmin(self.val_loss)
        logger.info("Best Metric: %d | val loss %.5f score %.2f",
                    best_bleu+1, self.val_loss[best_bleu],
                    self.val_metric[best_bleu])
        handle.write("Best Metric: %d | val loss %.5f score %.2f\n"
                     % (best_bleu+1, self.val_loss[best_bleu],
                        self.val_metric[best_bleu]))
        logger.info("Best loss: %d | val loss %.5f score %.2f",
                    best_loss+1, self.val_loss[best_loss],
                    self.val_metric[best_loss])
        handle.write("Best loss: %d | val loss %.5f score %.2f\n"
                     % (best_loss+1, self.val_loss[best_loss],
                        self.val_metric[best_loss]))
        logger.info("Early stopping marker: wait/patience: %d/%d\n",
                    self.wait, self.patience)
        handle.write("Early stopping marker: wait/patience: %d/%d\n" %
                     (self.wait, self.patience))
        handle.close() 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def test_nanfunctions_matrices_general():
    # Check that it works and that type and
    # shape are preserved
    # 2018-04-29: moved here from core.tests.test_nanfunctions
    mat = np.matrix(np.eye(3))
    for f in (np.nanargmin, np.nanargmax, np.nansum, np.nanprod,
              np.nanmean, np.nanvar, np.nanstd):
        res = f(mat, axis=0)
        assert_(isinstance(res, np.matrix))
        assert_(res.shape == (1, 3))
        res = f(mat, axis=1)
        assert_(isinstance(res, np.matrix))
        assert_(res.shape == (3, 1))
        res = f(mat)
        assert_(np.isscalar(res))

    for f in np.nancumsum, np.nancumprod:
        res = f(mat, axis=0)
        assert_(isinstance(res, np.matrix))
        assert_(res.shape == (3, 3))
        res = f(mat, axis=1)
        assert_(isinstance(res, np.matrix))
        assert_(res.shape == (3, 3))
        res = f(mat)
        assert_(isinstance(res, np.matrix))
        assert_(res.shape == (1, 3*3)) 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def test_nanargmax(self):
        tgt = np.argmax(self.mat)
        for mat in self.integer_arrays():
            assert_equal(np.nanargmax(mat), tgt) 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def frequency_at_max_power(self):
        """Returns the frequency corresponding to the highest peak in the periodogram."""
        return self.frequency[np.nanargmax(self.power)] 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def transit_time_at_max_power(self):
        """Returns the transit time corresponding to the highest peak in the periodogram."""
        return self.transit_time[np.nanargmax(self.power)] 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def duration_at_max_power(self):
        """Returns the duration corresponding to the highest peak in the periodogram."""
        return self.duration[np.nanargmax(self.power)] 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def depth_at_max_power(self):
        """Returns the depth corresponding to the highest peak in the periodogram."""
        return self.depth[np.nanargmax(self.power)] 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def test_nanargmax(self):
        def ref_impl(a):
            return np.nanargmax(a)

        def sdc_impl(a):
            return numpy_like.nanargmax(a)

        sdc_func = self.jit(sdc_impl)

        cases = [[np.nan, np.nan, np.inf, np.nan], [5, 2, -9, 333, -4], [3.3, 5.4, np.nan, 7.9]]
        for case in cases:
            a = np.array(case)
            with self.subTest(data=case):
                np.testing.assert_array_equal(sdc_func(a), ref_impl(a)) 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def discrete_planning(self, state: rlt.FeatureData) -> Tuple[int, np.ndarray]:
        # For discrete actions, we use random shoots to get the best next action
        random_action_seqs = list(
            itertools.product(range(self.action_dim), repeat=self.plan_horizon_length)
        )
        random_action_seqs = random.choices(random_action_seqs, k=self.cem_pop_size)
        action_solutions = torch.zeros(
            self.cem_pop_size, self.plan_horizon_length, self.action_dim
        )
        for i, action_seq in enumerate(random_action_seqs):
            for j, act_idx in enumerate(action_seq):
                action_solutions[i, j, act_idx] = 1
        acc_rewards = self.acc_rewards_of_all_solutions(state, action_solutions)

        first_action_tally = np.zeros(self.action_dim)
        reward_tally = np.zeros(self.action_dim)
        # pyre-fixme[6]: Expected `Iterable[Variable[_T2]]` for 2nd param but got
        #  `float`.
        for action_seq, acc_reward in zip(random_action_seqs, acc_rewards):
            first_action = action_seq[0]
            first_action_tally[first_action] += 1
            reward_tally[first_action] += acc_reward

        best_next_action_idx = np.nanargmax(reward_tally / first_action_tally)
        best_next_action_one_hot = torch.zeros(self.action_dim).float()
        best_next_action_one_hot[best_next_action_idx] = 1

        logger.debug(
            f"Choose action {best_next_action_idx}."
            f"Stats: {reward_tally} / {first_action_tally}"
            f" = {reward_tally/first_action_tally} "
        )
        return best_next_action_idx, best_next_action_one_hot 

# 需要导入模块: import numpy [as 别名]
# 或者: from numpy import nanargmax [as 别名]
def early_stop(metrics_hist, criterion, patience):
    if not np.all(np.isnan(metrics_hist[criterion])):
        if len(metrics_hist[criterion]) >= patience:
            if criterion == 'loss_dev': 
                return np.nanargmin(metrics_hist[criterion]) < len(metrics_hist[criterion]) - patience
            else:
                return np.nanargmax(metrics_hist[criterion]) < len(metrics_hist[criterion]) - patience
    else:
        #keep training if criterion results have all been nan so far
        return False