conversation.py 文件源码

def n_messages_chi_square(self, time_interval):
        """
        Computes a chi square test against the null hypothesis that the number
        of messages is uniformly distributed across the time interval. Only
        makes sense for the time intervals 'minute in hour', 'minute in day',
        'hour' since those ones have a fixed number of values.

        Args:
            time_interval: One of 'minute in hour', 'minute in day', 'hour'.

        Returns:
            chisq: A float representing the chi square statistic where the
                observations consist of the number of messages in each value of
                time_interval and the null hypothesis is that the number of
                messages is uniformly distributed.
            p: A float representing the p-value of the chi square test.
        """
        valid_time_intervals = ['minute in hour', 'minute in day', 'hour']
        if time_interval not in valid_time_intervals:
            raise ValueError('time_interval must be in {}'.format(valid_time_intervals))
        result = chisquare(self.get_n_messages_in_time_interval(time_interval))
        return (result.statistic, result.pvalue)