numpy.nanmean - python examples

Here are the examples of the python api numpy.nanmean taken from open source projects. By voting up you can indicate which examples are most useful and appropriate.

145 Examples 7

3 View Complete Implementation : table_extractor.py
Copyright MIT License
Author : UW-Deepdive-Infrastructure
def line_word_height(line):
    # For each line, get words
    words = line.find_all('span', 'ocrx_word')
    word_heights = []
    for word_idx, word in enumerate(words):
        wordbbox = helpers.extractbbox(word.get('satle'))
        word_heights.append(wordbbox['y2'] - wordbbox['y1'])

    avg = 0 if len(words) == 0 else np.nanmean(word_heights)

    return avg

3 View Complete Implementation : impala_distributed_dmlab.py
Copyright Apache License 2.0
Author : rlgraph
def _calc_mean_return(records):
    size = records[3]["rewards"].size
    rewards = records[3]["rewards"].reshape((size,))
    terminals = records[3]["terminals"].reshape((size,))
    returns = list()
    return_ = 0.0
    for r, t in zip(rewards, terminals):
        return_ += r
        if t:
            returns.append(return_)
            return_ = 0.0

    return np.nanmean(returns)

3 View Complete Implementation : area_stats.py
Copyright MIT License
Author : UW-Deepdive-Infrastructure
def summarizeDocameent(area_stats):
    # Don't use areas with 1 line or no words in creating summary statistics
    return {
        'word_separation_mean': np.nanmean([np.nanmean(area['word_distances']) for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_separation_median': np.nanmedian([np.nanmean(area['word_distances']) for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_separation_std': np.nanstd([np.nanmean(area['word_distances'])for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_separation_index_mean': np.nanmean([area['word_separation_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_separation_index_median': np.nanmedian([area['word_separation_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_separation_index_std': np.nanstd([area['word_separation_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_index_mean': np.nanmean([area['word_height_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_index_median': np.nanmedian([area['word_height_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_index_std': np.nanstd([area['word_height_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_area_index_mean': np.nanmean([area['word_area_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_area_index_median': np.nanmedian([area['word_area_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_area_index_std': np.nanstd([area['word_area_index'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_avg': np.nanmean([area['word_height_avg'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_avg_median': np.nanmedian([area['word_height_avg'] for area in area_stats if area['words'] > 0 and area['lines'] > 1]),
        'word_height_avg_std': np.nanstd([area['word_height_avg'] for area in area_stats if area['words'] > 0 and area['lines'] > 1])
    }

3 View Complete Implementation : impala_cartpole.py
Copyright Apache License 2.0
Author : rlgraph
def _calc_mean_return(records):
    size = records[3]["rewards"].size
    rewards = records[3]["rewards"].reshape((size,))
    terminals = records[3]["terminals"].reshape((size,))
    returns = list()
    return_ = 0.0
    for r, t in zip(rewards, terminals):
        return_ += r
        if t:
            returns.append(return_)
            return_ = 0.0

    return np.nanmean(returns)

3 View Complete Implementation : impala_openai_gym_with_lstm.py
Copyright Apache License 2.0
Author : rlgraph
def _calc_mean_return(records):
    size = records[3]["rewards"].size
    rewards = records[3]["rewards"].reshape((size,))
    terminals = records[3]["terminals"].reshape((size,))
    returns = list()
    return_ = 0.0
    for r, t in zip(rewards, terminals):
        return_ += r
        if t:
            returns.append(return_)
            return_ = 0.0

    return np.nanmean(returns)

3 View Complete Implementation : test_utils.py
Copyright GNU General Public License v3.0
Author : NVISO-BE
    def test_decision_frontier_mad_zero(self):
        values_array = [1, 1]
        sensitivity = 10
        # Here mad = 0
        # mad formula:
        # mad = np.nanmedian(np.absolute(values_array - np.nanmedian(values_array, 0)), 0)

        res = helpers.utils.get_decision_frontier("mad", values_array, sensitivity, "low")
        # So use std:
        expected_value = np.nanmean(values_array) - sensitivity * np.std(values_array)
        self.astertEqual(res, expected_value)

        res = helpers.utils.get_decision_frontier("mad", values_array, sensitivity, "high")
        # So use std:
        expected_value = np.nanmean(values_array) + sensitivity * np.std(values_array)
        self.astertEqual(res, expected_value)

3 View Complete Implementation : test_utils.py
Copyright GNU General Public License v3.0
Author : NVISO-BE
    def test_decision_frontier_madpos_zero(self):
        values_array = [1, 1]
        sensitivity = 10
        # Here mad = 0
        # mad formula:
        # mad = np.nanmedian(np.absolute(values_array - np.nanmedian(values_array, 0)), 0)

        res = helpers.utils.get_decision_frontier("madpos", values_array, sensitivity, "low")
        # So use std:
        expected_value = np.nanmean(values_array) - sensitivity * np.std(values_array)
        expected_value = np.float64(max([expected_value, 0]))
        self.astertEqual(res, expected_value)

        res = helpers.utils.get_decision_frontier("madpos", values_array, sensitivity, "high")
        # So use std:
        expected_value = np.nanmean(values_array) + sensitivity * np.std(values_array)
        expected_value = np.float64(max([expected_value, 0]))
        self.astertEqual(res, expected_value)

3 View Complete Implementation : test_utils.py
Copyright GNU General Public License v3.0
Author : NVISO-BE
    def test_decision_frontier_stdev_high(self):
        for values_array in list_values_array:
            nanmean_values_array = np.nanmean(values_array)
            std_values_array = np.std(values_array)
            for sensitivity in list_sensitivity:
                expected_res = nanmean_values_array + sensitivity * std_values_array
                if expected_res < 0:
                    with self.astertLogs(logging.logger, level='DEBUG'):
                        res = helpers.utils.get_decision_frontier("stdev", values_array, sensitivity, "high")
                else:
                    res = helpers.utils.get_decision_frontier("stdev", values_array, sensitivity, "high")
                self.astertEqual(res, expected_res)

3 View Complete Implementation : early_stopping.py
Copyright MIT License
Author : zalando
def make_group_sequential(spending_function='obrien_fleming', estimated_sample_size=None, alpha=0.05, cap=8):
    """ A closure to the group_sequential function. """
    def go(x, y, x_denominators=1, y_denominators=1):

        # these next too lines are wrong, but they are bug-compatible with v0.6.13 !
        x = x / np.nanmean(x_denominators)
        y = y / np.nanmean(y_denominators)

        return group_sequential(x, y, spending_function, estimated_sample_size, alpha, cap)
    return go

3 View Complete Implementation : test_utils.py
Copyright GNU General Public License v3.0
Author : NVISO-BE
    def test_decision_frontier_stdev_low(self):
        for values_array in list_values_array:
            nanmean_values_array = np.nanmean(values_array)
            std_values_array = np.std(values_array)
            for sensitivity in list_sensitivity:

                expected_res = nanmean_values_array - sensitivity * std_values_array
                if expected_res < 0:
                    with self.astertLogs(logging.logger, level='DEBUG'):
                        res = helpers.utils.get_decision_frontier("stdev", values_array, sensitivity, "low")
                else:
                    res = helpers.utils.get_decision_frontier("stdev", values_array, sensitivity, "low")

                self.astertEqual(res, expected_res)