python类TimeGrouper()的实例源码

def quiet_days(series,
               n_days=5,
               period='min'):
    """
    Given a pandas time series *series*, return a mapping between
    months and the *n_days* quietest days for that month. See (4) of
    Love and Gannon. The parameter *period* gives the sampling period
    (`'min'` = 1 measurement per minute).
    """
    quiet_day_map = {}
    delta_H_i = series.diff().abs().\
                groupby(PD.TimeGrouper(freq='D')).\
                filter(lambda x: x.isnull().sum() <= int(0.5 * SAMPLES[period])).\
                groupby(PD.TimeGrouper(freq='D')).mean()
    for month, delta_H_i_month in delta_H_i.groupby(PD.TimeGrouper(freq='M')):
        quiet_day_map[month] = delta_H_i_month.nsmallest(n_days).sort_index().index
    return quiet_day_map

def chebyshev_fit(series,
                  quiet_day_map,
                  deg=10):
    """
    ???
    """
    grouped_by_day = series.groupby(PD.TimeGrouper(freq='D'))
    x = []
    y = []
    for month_end in sorted(quiet_day_map):
        for quiet_day in quiet_day_map[month_end]:
            quiet_day_H = grouped_by_day.get_group(quiet_day)
            quiet_day_H = quiet_day_H[quiet_day_H.notnull()]
            x.extend([toJ2000(time_stamp.to_datetime()) for time_stamp in quiet_day_H.index])
            y.extend(quiet_day_H.values)
    return NP.polynomial.chebyshev.chebfit(x, y, deg, full=True)

def test_dates(self):
        for t in self.mock_tweets:
            self.feels_db.insert_tweet(t)
        self.assertEqual(len(self.feels_db.tweet_dates), 3)

        tweets = []
        with open(self.tweets_data_path) as tweets_file:
            lines = filter(None, (line.rstrip() for line in tweets_file))
            for line in lines:
                try:
                    tweets.append(Tweet(json.loads(line)))
                except KeyError:
                    pass
        for t in tweets:
            self.feels_db.insert_tweet(t)
        self.assertEqual(len(self.feels_db.tweet_dates), 105)
        df = self.feels_db.tweet_dates
        timebox = timedelta(seconds=60)
        second = timedelta(seconds=1)
        df = df.groupby(pd.TimeGrouper(freq=f'{int(timebox/second)}S')).size()
        df = df[df != 0]
        print(df)
        self.assertEqual(len(df), 3)
        self.assertEqual(df.iloc[0], 103)

def test_resample_basic(self):
        rng = date_range('1/1/2000 00:00:00', '1/1/2000 00:13:00', freq='min',
                         name='index')
        s = Series(np.random.randn(14), index=rng)
        result = s.resample('5min', closed='right', label='right').mean()

        exp_idx = date_range('1/1/2000', periods=4, freq='5min', name='index')
        expected = Series([s[0], s[1:6].mean(), s[6:11].mean(), s[11:].mean()],
                          index=exp_idx)
        assert_series_equal(result, expected)
        self.assertEqual(result.index.name, 'index')

        result = s.resample('5min', closed='left', label='right').mean()

        exp_idx = date_range('1/1/2000 00:05', periods=3, freq='5min',
                             name='index')
        expected = Series([s[:5].mean(), s[5:10].mean(),
                           s[10:].mean()], index=exp_idx)
        assert_series_equal(result, expected)

        s = self.series
        result = s.resample('5Min').last()
        grouper = TimeGrouper(Minute(5), closed='left', label='left')
        expect = s.groupby(grouper).agg(lambda x: x[-1])
        assert_series_equal(result, expect)

def test_resample_frame_basic(self):
        df = tm.makeTimeDataFrame()

        b = TimeGrouper('M')
        g = df.groupby(b)

        # check all cython functions work
        funcs = ['add', 'mean', 'prod', 'min', 'max', 'var']
        for f in funcs:
            g._cython_agg_general(f)

        result = df.resample('A').mean()
        assert_series_equal(result['A'], df['A'].resample('A').mean())

        result = df.resample('M').mean()
        assert_series_equal(result['A'], df['A'].resample('M').mean())

        df.resample('M', kind='period').mean()
        df.resample('W-WED', kind='period').mean()

def test_resample_ohlc(self):
        s = self.series

        grouper = TimeGrouper(Minute(5))
        expect = s.groupby(grouper).agg(lambda x: x[-1])
        result = s.resample('5Min').ohlc()

        self.assertEqual(len(result), len(expect))
        self.assertEqual(len(result.columns), 4)

        xs = result.iloc[-2]
        self.assertEqual(xs['open'], s[-6])
        self.assertEqual(xs['high'], s[-6:-1].max())
        self.assertEqual(xs['low'], s[-6:-1].min())
        self.assertEqual(xs['close'], s[-2])

        xs = result.iloc[0]
        self.assertEqual(xs['open'], s[0])
        self.assertEqual(xs['high'], s[:5].max())
        self.assertEqual(xs['low'], s[:5].min())
        self.assertEqual(xs['close'], s[4])

def test_apply_iteration(self):
        # #2300
        N = 1000
        ind = pd.date_range(start="2000-01-01", freq="D", periods=N)
        df = DataFrame({'open': 1, 'close': 2}, index=ind)
        tg = TimeGrouper('M')

        _, grouper, _ = tg._get_grouper(df)

        # Errors
        grouped = df.groupby(grouper, group_keys=False)
        f = lambda df: df['close'] / df['open']

        # it works!
        result = grouped.apply(f)
        self.assertTrue(result.index.equals(df.index))

def test_panel_aggregation(self):
        ind = pd.date_range('1/1/2000', periods=100)
        data = np.random.randn(2, len(ind), 4)
        wp = pd.Panel(data, items=['Item1', 'Item2'], major_axis=ind,
                      minor_axis=['A', 'B', 'C', 'D'])

        tg = TimeGrouper('M', axis=1)
        _, grouper, _ = tg._get_grouper(wp)
        bingrouped = wp.groupby(grouper)
        binagg = bingrouped.mean()

        def f(x):
            assert (isinstance(x, Panel))
            return x.mean(1)

        result = bingrouped.agg(f)
        tm.assert_panel_equal(result, binagg)

def test_fails_on_no_datetime_index(self):
        index_names = ('Int64Index', 'Index', 'Float64Index', 'MultiIndex')
        index_funcs = (tm.makeIntIndex,
                       tm.makeUnicodeIndex, tm.makeFloatIndex,
                       lambda m: tm.makeCustomIndex(m, 2))
        n = 2
        for name, func in zip(index_names, index_funcs):
            index = func(n)
            df = DataFrame({'a': np.random.randn(n)}, index=index)
            with tm.assertRaisesRegexp(TypeError,
                                       "Only valid with DatetimeIndex, "
                                       "TimedeltaIndex or PeriodIndex, "
                                       "but got an instance of %r" % name):
                df.groupby(TimeGrouper('D'))

        # PeriodIndex gives a specific error message
        df = DataFrame({'a': np.random.randn(n)}, index=tm.makePeriodIndex(n))
        with tm.assertRaisesRegexp(TypeError,
                                   "axis must be a DatetimeIndex, but "
                                   "got an instance of 'PeriodIndex'"):
            df.groupby(TimeGrouper('D'))

def get_ts_data():
    requested_facts = text_input.value.split(",")[:8]
    requested_facts = [f.strip() for f in requested_facts]
    req_df = preprocessing.get_facts_from_list(df, requested_facts)
    ts = preprocessing.make_timeseries(req_df)
    ts.columns = ts.columns.droplevel(0)
    ts = ts.groupby(pd.TimeGrouper(freq=timegroupoptionsmapper[timegroup.active])).sum()
    return ts

def get_daily_gain(data, timezone=pytz.timezone('US/Pacific')):
    """Obtain the daily gain.

    Attributes
    ----------
    data: list
        The list of dictionaries where each dictionary gives information on one
        completed submission. It is the output of the `get_data` function

    timezone: pytz.timezone
        A valid pytz timezone. Default is the Pacific Standard Time, which is
        the one used by Udacity

    Returns
    -------
        A Pandas Series where the indices are the days and the values are the
        total gain in USD of that day. The time zone is the Pacific Standard
        Time.
    """

    date_price_data = np.array([(d['completed_at'], d['price']) for d in data])

    price_series = pd.Series(date_price_data[:, 1].astype(float),
                             index=pd.to_datetime(date_price_data[:, 0]))
    price_series = price_series.sort_index()

    # Convert timezone
    utc = pytz.utc
    price_series = price_series.tz_localize(utc).tz_convert(timezone)


    # Calculate the gain by day
    daily_gain = price_series.groupby(pd.TimeGrouper('D')).sum()

    return daily_gain

def test_custom_grouper(self):

        dti = DatetimeIndex(freq='Min', start=datetime(2005, 1, 1),
                            end=datetime(2005, 1, 10))

        s = Series(np.array([1] * len(dti)), index=dti, dtype='int64')

        b = TimeGrouper(Minute(5))
        g = s.groupby(b)

        # check all cython functions work
        funcs = ['add', 'mean', 'prod', 'ohlc', 'min', 'max', 'var']
        for f in funcs:
            g._cython_agg_general(f)

        b = TimeGrouper(Minute(5), closed='right', label='right')
        g = s.groupby(b)
        # check all cython functions work
        funcs = ['add', 'mean', 'prod', 'ohlc', 'min', 'max', 'var']
        for f in funcs:
            g._cython_agg_general(f)

        self.assertEqual(g.ngroups, 2593)
        self.assertTrue(notnull(g.mean()).all())

        # construct expected val
        arr = [1] + [5] * 2592
        idx = dti[0:-1:5]
        idx = idx.append(dti[-1:])
        expect = Series(arr, index=idx)

        # GH2763 - return in put dtype if we can
        result = g.agg(np.sum)
        assert_series_equal(result, expect)

        df = DataFrame(np.random.rand(len(dti), 10),
                       index=dti, dtype='float64')
        r = df.groupby(b).agg(np.sum)

        self.assertEqual(len(r.columns), 10)
        self.assertEqual(len(r.index), 2593)

def test_resample_nunique(self):

        # GH 12352
        df = DataFrame({
            'ID': {pd.Timestamp('2015-06-05 00:00:00'): '0010100903',
                   pd.Timestamp('2015-06-08 00:00:00'): '0010150847'},
            'DATE': {pd.Timestamp('2015-06-05 00:00:00'): '2015-06-05',
                     pd.Timestamp('2015-06-08 00:00:00'): '2015-06-08'}})
        r = df.resample('D')
        g = df.groupby(pd.Grouper(freq='D'))
        expected = df.groupby(pd.TimeGrouper('D')).ID.apply(lambda x:
                                                            x.nunique())
        self.assertEqual(expected.name, 'ID')

        for t in [r, g]:
            result = r.ID.nunique()
            assert_series_equal(result, expected)

        # TODO
        # this should have name
        # https://github.com/pydata/pandas/issues/12363
        expected.name = None
        result = df.ID.resample('D').nunique()
        assert_series_equal(result, expected)

        result = df.ID.groupby(pd.Grouper(freq='D')).nunique()
        assert_series_equal(result, expected)

def test_count(self):
        self.ts[::3] = np.nan

        expected = self.ts.groupby(lambda x: x.year).count()

        grouper = TimeGrouper('A', label='right', closed='right')
        result = self.ts.groupby(grouper).count()
        expected.index = result.index
        assert_series_equal(result, expected)

        result = self.ts.resample('A').count()
        expected.index = result.index
        assert_series_equal(result, expected)

def test_groupby_with_empty(self):
        index = pd.DatetimeIndex(())
        data = ()
        series = pd.Series(data, index)
        grouper = pd.tseries.resample.TimeGrouper('D')
        grouped = series.groupby(grouper)
        assert next(iter(grouped), None) is None

def test_groupby_with_timegrouper(self):
        # GH 4161
        # TimeGrouper requires a sorted index
        # also verifies that the resultant index has the correct name
        import datetime as DT
        df_original = DataFrame({
            'Buyer': 'Carl Carl Carl Carl Joe Carl'.split(),
            'Quantity': [18, 3, 5, 1, 9, 3],
            'Date': [
                DT.datetime(2013, 9, 1, 13, 0),
                DT.datetime(2013, 9, 1, 13, 5),
                DT.datetime(2013, 10, 1, 20, 0),
                DT.datetime(2013, 10, 3, 10, 0),
                DT.datetime(2013, 12, 2, 12, 0),
                DT.datetime(2013, 9, 2, 14, 0),
            ]
        })

        # GH 6908 change target column's order
        df_reordered = df_original.sort_values(by='Quantity')

        for df in [df_original, df_reordered]:
            df = df.set_index(['Date'])

            expected = DataFrame(
                {'Quantity': np.nan},
                index=date_range('20130901 13:00:00',
                                 '20131205 13:00:00', freq='5D',
                                 name='Date', closed='left'))
            expected.iloc[[0, 6, 18], 0] = np.array(
                [24., 6., 9.], dtype='float64')

            result1 = df.resample('5D') .sum()
            assert_frame_equal(result1, expected)

            df_sorted = df.sort_index()
            result2 = df_sorted.groupby(pd.TimeGrouper(freq='5D')).sum()
            assert_frame_equal(result2, expected)

            result3 = df.groupby(pd.TimeGrouper(freq='5D')).sum()
            assert_frame_equal(result3, expected)

def test_groupby_with_timegrouper_methods(self):
        # GH 3881
        # make sure API of timegrouper conforms

        import datetime as DT
        df_original = pd.DataFrame({
            'Branch': 'A A A A A B'.split(),
            'Buyer': 'Carl Mark Carl Joe Joe Carl'.split(),
            'Quantity': [1, 3, 5, 8, 9, 3],
            'Date': [
                DT.datetime(2013, 1, 1, 13, 0),
                DT.datetime(2013, 1, 1, 13, 5),
                DT.datetime(2013, 10, 1, 20, 0),
                DT.datetime(2013, 10, 2, 10, 0),
                DT.datetime(2013, 12, 2, 12, 0),
                DT.datetime(2013, 12, 2, 14, 0),
            ]
        })

        df_sorted = df_original.sort_values(by='Quantity', ascending=False)

        for df in [df_original, df_sorted]:
            df = df.set_index('Date', drop=False)
            g = df.groupby(pd.TimeGrouper('6M'))
            self.assertTrue(g.group_keys)
            self.assertTrue(isinstance(g.grouper, pd.core.groupby.BinGrouper))
            groups = g.groups
            self.assertTrue(isinstance(groups, dict))
            self.assertTrue(len(groups) == 3)

def backtest(self, data_frame):
        dfs = data_frame.groupby(pd.TimeGrouper(freq='D'))
        # only choose trading days
        dfs = [(d, df) for (d, df) in dfs if df.shape[0]]

        if sys.version_info[0] < 3:
            exit_dfs = [self._compute_daily_performance(daily_data)
                        for daily_data in dfs]
        else:
            exit_dfs = Parallel(n_jobs=-1, verbose=50)(
                delayed(self._compute_daily_performance)(daily_data)
                for daily_data in dfs)

        return pd.concat(exit_dfs)

def timeGroup (frame, step, file_csv): 
    dfx = frame.set_index(['time'])
    dfx = dfx[dfx.status == 'accepted']
    grouper = dfx.groupby([pd.TimeGrouper(step), 'Miner'])
    dfTimeS = grouper['Miner'].count().unstack('Miner').fillna(0)
    dfTimeS = dfTimeS[:-1] #There are arguments for and against this. The final group is not 'complete', so we will not include it.
    dfTimeS.to_csv(file_csv, encoding='utf-8')
    return dfTimeS

#Summary tables

def calculate_latest_coeffs(self):
        unit_topic_tmpl = "{campus}/{building}/{unit}/{point}"
        unit_points = [self.power_name]
        df = None

        #Get data
        unit = self.temp_unit
        for point in unit_points:
            if point == self.power_name:
                unit = self.power_unit
            unit_topic = unit_topic_tmpl.format(campus=self.site,
                                                building=self.building,
                                                unit=unit,
                                                point=point)
            result = self.vip.rpc.call('platform.historian',
                                       'query',
                                       topic=unit_topic,
                                       count=self.no_of_recs_needed,
                                       order="LAST_TO_FIRST").get(timeout=10000)
            df2 = pd.DataFrame(result['values'], columns=[self.ts_name, point])
            df2[self.ts_name] = pd.to_datetime(df2[self.ts_name])
            df2 = df2.groupby([pd.TimeGrouper(key=self.ts_name, freq=self.aggregate_freq)]).mean()
            # df2[self.ts_name] = df2[self.ts_name].apply(lambda dt: dt.replace(second=0, microsecond=0))
            df = df2 if df is None else pd.merge(df, df2, how='outer', left_index=True, right_index=True)

        #Calculate coefficients
        result_df = self.calculate_coeffs(df)


        # Publish coeffs to store
        #if coeffs is not None:
        #    self.save_coeffs(coeffs, subdevice)

def calculate_latest_coeffs(self):
        unit_topic_tmpl = "{campus}/{building}/{unit}/{point}"
        unit_points = [self.fan_power_name, self.static_pressure_name, self.air_flow_rate_name]
        df = None

        for point in unit_points:
            unit_topic = unit_topic_tmpl.format(campus=self.site,
                                                building=self.building,
                                                unit=self.unit,
                                                point=point)
            result = self.vip.rpc.call('platform.historian',
                                       'query',
                                       topic=unit_topic,
                                       count=self.no_of_recs_needed,
                                       order="LAST_TO_FIRST").get(timeout=1000)
            df2 = pd.DataFrame(result['values'], columns=[self.ts_name, point])
            self.convert_units_to_SI(df2, point, result['metadata']['units'])
            df2[self.ts_name] = pd.to_datetime(df2[self.ts_name])
            df2 = df2.groupby([pd.TimeGrouper(key=self.ts_name, freq=self.aggregate_freq)]).mean()
            #df2[self.ts_name] = df2[self.ts_name].apply(lambda dt: dt.replace(second=0, microsecond=0))
            df = df2 if df is None else pd.merge(df, df2, how='outer', left_index=True, right_index=True)

        #print(df)
        coeffs = self.calculate_coeffs(df)
        # Publish coeffs to store
        if coeffs is not None:
            self.save_coeffs(coeffs)

def Calculate_Liquidity_Coeff(df_list):
    #list of lists of liq coeff per bond
    liq_arr_list = []

    #print 'df_list size: ', len(df_list)
    for df in df_list:
        if df.empty:
            continue

        # A temporary array for holding liquidity beta for each month
        liq_arr = [np.nan] * num_months_CONST
        #print df['cusip_id'][0]

        # Group dataframe on index by month
        for date, df_group in df.groupby(pd.TimeGrouper("M")):
            month = ''.join([str(date.month),str(date.year)])
            month_key = month_keys[month]

            # When there are some data in current month,
            if df_group.shape[0] > 0:
                # Run regression (as equation (2)) to get liquidity measure
                y,X = dmatrices('excess_return_1 ~ yld_pt + volume_and_sign', 
                                data=df_group, return_type='dataframe')
                #print date, X.shape
                mod = sm.OLS(y,X)
                res = mod.fit()

                #set specific months with liquidity factors
                #res.params(2) = liquidity coefficient
                liq_arr[month_key] = res.params[2]

        liq_arr_list.append(liq_arr)    #store all liq coeff for each month per bond

    return liq_arr_list

def wip_chart(cfd_data, frequency="1W-MON", start_column=None, end_column=None, title=None, ax=None):
    if len(cfd_data.index) == 0:
        raise UnchartableData("Cannot draw WIP chart with no data")

    if start_column is None:
        start_column = cfd_data.columns[1]
    if end_column is None:
        end_column = cfd_data.columns[-1]

    if ax is None:
        fig, ax = plt.subplots()

    if title is not None:
        ax.set_title(title)

    wip_data = pd.DataFrame({'wip': cfd_data[start_column] - cfd_data[end_column]})

    groups = wip_data[['wip']].groupby(pd.TimeGrouper(frequency, label='left'))
    labels = [x[0].strftime("%d/%m/%Y") for x in groups]

    groups.boxplot(subplots=False, ax=ax, showmeans=True, return_type='axes')
    ax.set_xticklabels(labels, rotation=70, size='small')

    ax.set_xlabel("Week")
    ax.set_ylabel("WIP")

    return ax

def get_savings(data, column_debit='Debit', column_credit='Credit', dt_start=None, dt_end=None, aggregation_period='M'):
    """ Consumes the checking account data and returns the monthly savings rate.

    Args:
        data (dataframe): The panadas dataframe containing at least a debit and a credit column.
        column_debit (str): The column name for the debit column.
        column_credit (str): The column name for the credit column.
        dt_start (str): The start date (specific if given '2012-11-11' or the month '2012-11')
            from were the savings should be calculated.
        dt_end (str): The end date (specific if given '2012-11-11' or the month '2012-11')
            to were the savings should be calculated.
        aggregation_period (str): Single string character like 'M' for month specifying, over which period the savings
            are aggregated. A full specification can be found here:
            http://pandas.pydata.org/pandas-docs/stable/timeseries.html#timeseries-offset-aliases

    Returns:
        A pandas data frame, with an additional 'Savings' column and the time difference between start and end
        represented with a single row for each aggregation interval that is not null.

    """
    if not isinstance(data.index, pd.DatetimeIndex):
        logging.getLogger().error("A pandas datetimeindex is required for the given dataframe")
        return pd.DataFrame()

    # create a copy of the indexed original data frame
    aggregated = data[dt_start:dt_end][[column_debit, column_credit]].copy()

    aggregated = aggregated.groupby(pd.TimeGrouper(aggregation_period)).sum()
    aggregated = aggregated.fillna(0)

    aggregated['Savings'] = aggregated[column_credit] - aggregated[column_debit]
    return aggregated

def test_basicStats(self):
        filepath =  RESOURCE_PATH + "\\unittest\\test_hb_basic01.csv"
        data = utils.loadIntradayData(filepath).set_index('datetime')
        stats = hbStats.groupByBasicStats(pd.TimeGrouper(freq='d'), data)

        self.assertEqual(stats.iloc[0]['count'], 16)
        self.assertEqual(stats.iloc[0]['max'], 70)
        self.assertEqual(stats.iloc[0]['min'], 50)
        self.assertEqual(stats.iloc[0]['mean'], 60)

def plotYearMonthStatsHb(data):
    #pd.groupby(b,by=[b.index.month,b.index.year])
    data.groupby(pd.TimeGrouper(freq='M')).mean().plot()
    sns.plt.show()

def avg_wl(self, numObs=50, avgtype='stdWL', grptype='bytime', grper='12M'):
        """Calculates standardized statistics for a list of stations or a huc from the USGS

        :param numObs: minimum observations per site required to include site in analysis; default is 50
        :param avgtype: averaging technique for site data; options are 'avgDiffWL' and 'stdWL'; default is 'stWL'
        :param grptype: way to group the averaged data; options are 'bytime' or 'monthly' or user input; default 'bytime'
        :param grper: only used if 'bytime' called; defaults to '12M'; other times can be put in
        :return:
        """

        data = self.cleanGWL(self.data)
        # stationWL = pd.merge(siteinfo, data, on = 'site_no')
        data.reset_index(inplace=True)
        data.set_index(['datetime'], inplace=True)
        # get averages by year, month, and site number
        site_size = data.groupby('site_no').size()
        wl_long = data[data['site_no'].isin(list(site_size[site_size >= numObs].index.values))]
        siteList = list(wl_long.site_no.unique())
        for site in siteList:
            mean = wl_long.ix[wl_long.site_no == site, 'value'].mean()
            std = wl_long.ix[wl_long.site_no == site, 'value'].std()
            wl_long.ix[wl_long.site_no == site, 'avgDiffWL'] = wl_long.ix[wl_long.site_no == site, 'value'] - mean
            wl_long.ix[wl_long.site_no == site, 'stdWL'] = wl_long.ix[wl_long.site_no == site, 'avgDiffWL'] / std

        if grptype == 'bytime':
            grp = pd.TimeGrouper(grper)
        elif grptype == 'monthly':
            grp = wl_long.index.month
        else:
            grp = grptype
        wl_stats = wl_long.groupby([grp])[avgtype].agg({'mean': np.mean, 'median': np.median,
                                                        'standard': np.std,
                                                        'cnt': (lambda x: np.count_nonzero(~np.isnan(x))),
                                                        'err_pls': (lambda x: np.mean(x) + (np.std(x) * 1.96)),
                                                        'err_min': (lambda x: np.mean(x) - (np.std(x) * 1.96))})

        return wl_stats

def fetchbin(self, start=None, end=None, binsize=timedelta(seconds=60),
                 empty=False):
        """
        Returns a generator that can be used to iterate over the tweet data
        based on ``binsize``.

        :param start: Query start date.
        :type start: datetime
        :param end: Query end date.
        :type end: datetime
        :param binsize: Time duration for each bin for tweet grouping.
        :type binsize: timedelta
        :param empty: Determines whether empty dataframes will be yielded.
        :type empty: boolean
        :returns: A dataframe along with time boundaries for the data.
        :rtype: tuple
        """
        second = timedelta(seconds=1)
        if start is None: start=self.start-second
        if end is None: end=self.end
        if start == self.start: start = start-second
        df = self.tweet_dates
        df = df.groupby(pd.TimeGrouper(freq=f'{int(binsize/second)}S')).size()
        df = df[df.index > start - binsize]
        if not empty: df = df[df != 0]
        conn = sqlite3.connect(self._db, detect_types=sqlite3.PARSE_DECLTYPES)
        c = conn.cursor()
        c.execute(
            "SELECT * FROM tweets WHERE created_at > ? AND created_at <= ?",
            (start, end)
            )
        for i in range(0,len(df)):
            frame = []
            if df.iloc[i] > 0:
                frame = pd.DataFrame.from_records(
                    data=c.fetchmany(df.iloc[i]), columns=self.fields,
                    index='created_at'
                    )
            left = df.index[i].to_pydatetime()
            right = left + binsize
            if len(frame)>0 or empty: yield TweetBin(frame, left, right)
        c.close()

def update(attrname, old, new):
    subset = get_subset(dictchooser.value)
    new_absolute_source = subset[0] \
                                        .ix[:, :top_n.value] \
                                        .groupby(pd.TimeGrouper(freq=timegroupoptionsmapper[timegroup.active])) \
                                        .sum().fillna(0)
    new_relative_source = subset[1] \
                                        .ix[:, :top_n.value] \
                                        .groupby(pd.TimeGrouper(freq=timegroupoptionsmapper[timegroup.active])) \
                                        .sum().fillna(0)

    for old, new in zip(abs_arrangement, new_absolute_source.columns.tolist()):
        old.title.text = new
    for old, new in zip(rel_arrangement, new_relative_source.columns.tolist()):
        old.title.text = new

    new_abs_sources = [ColumnDataSource(dict(date=new_absolute_source.index,
                                             y=new_absolute_source[l]))
                                                for l in new_absolute_source.columns.tolist()]
    new_rel_sources = [ColumnDataSource(dict(date=new_relative_source.index,
                                             y=new_relative_source[l]))
                                                for l in new_relative_source.columns.tolist()]
    for old, new in zip(abs_sources, new_abs_sources):
        old.data.update(new.data)
    for old, new in zip(rel_sources, new_rel_sources):
        old.data.update(new.data)

    new_abs_point_sources = [ColumnDataSource(dict(date=[new_absolute_source[l].idxmax()],
                                               y=[new_absolute_source[l].max()],
                                               text=[str(int(new_absolute_source[l].max()))]
                                               )
                                          )
                                          for l in new_absolute_source.columns.tolist()]
    new_rel_point_sources = [ColumnDataSource(dict(date=[new_relative_source[l].idxmax()],
                                               y=[new_relative_source[l].max()],
                                               text=[str(int(new_relative_source[l].max()))]
                                               )
                                          )
                                          for l in new_relative_source.columns.tolist()]
    for old, new in zip(abs_point_sources, new_abs_point_sources):
        old.data.update(new.data)
    for old, new in zip(rel_point_sources, new_rel_point_sources):
        old.data.update(new.data)

# Create Input controls

def run(tags, time_span, step):
    data_sets = [Data(time_span, step, t) for t in tags]
    data_frames = []

    step_fmt = period_map.get(step)

    for ss in data_sets:
        if not ss.df.empty:
            ss.df['start_time'] = pd.to_datetime(ss.df['start'])
            ss.df['end_time'] = pd.to_datetime(ss.df['end'])

            ss.df.drop('start', axis=1, inplace=True)  # Drop `start` column.
            ss.df.drop('end', axis=1, inplace=True)  # Drop `end` column.
            ss.df.drop('tags', axis=1, inplace=True)  # Drop `tags` column.

            ss.df['duration'] = (ss.df['end_time'] - ss.df['start_time'])
            ss.df['duration'] = (
                (ss.df['duration'] / np.timedelta64(1, 's')) / 60
            )
            ss.df['interval'] = ss.df.end_time.dt.to_period(step_fmt)
            ss.df = ss.df.set_index('interval')  # `interval` column as index.

            ss.df.drop('start_time', axis=1, inplace=True)  # Drop `start_time`.
            ss.df.drop('end_time', axis=1, inplace=True)  # Drop `end_time`.

            ss.df = ss.df.groupby(
                pd.TimeGrouper(step_fmt),
                level=0,
            ).aggregate(
                np.sum
            )
            ss.df.rename(columns={'duration': ss.tag}, inplace=True)

            data_frames.append(ss.df)

    result = pd.concat(data_frames, axis=1)

    if step_fmt == 'D':
        result = result.to_timestamp()  # `PeriodIndex` to `DatetimeIndex`.
        result = result.asfreq('D', fill_value=0)  # Fill missing days.

    plot = result.plot(kind='bar')
    plot.set_title('Minutes spent by {p}'.format(p=step))
    plot.set_xlabel('{p}s'.format(p=step))
    plot.set_ylabel('minutes')

    plt.show()