finished boxplot

.gitignore

@@ -5,4 +5,5 @@
 opengrid/notebooks/index.html
 dist/
 build/
-opengrid.egg-info/
+opengrid.egg-info/
+.DS_Store

opengrid/library/analysis.py

@@ -9,6 +9,7 @@
 import datetime as dt
 import pandas as pd
 import numpy as np
+import numbers
 from opengrid.library.exceptions import EmptyDataFrame
 
 
@@ -68,22 +69,80 @@ def do_analysis(self, agg, starttime=dt.time.min, endtime=dt.time.max):
             self.result = pd.DataFrame()
 
 
-def standby(df, resolution='d'):
+def standby(df, resolution='24h', time_window=None):
     """
     Compute standby power
 
     Parameters
     ----------
     df : Pandas DataFrame
         Electricity Power
-    resolution : str
+    resolution : str, default='d'
+        Resolution of the computation.  Data will be resampled to this resolution (as mean) before computation
+        of the minimum.
+        String that can be parsed by the pandas resample function, example ='h', '15min', '6h'
+    time_window : tuple with start-hour and end-hour, default=None
+        Specify the start-time and end-time for the analysis.
+        Only data within this time window will be considered.
+        Both times have to be specified as string ('01:00', '06:30') or as datetime.time() objects
+
+    Returns
+    -------
+    df : pandas.Series with DateTimeIndex in the given resolution
     """
 
+    def parse_time(t):
+        if isinstance(t, numbers.Number):
+            return pd.Timestamp.utcfromtimestamp(t).time()
+        else:
+            return pd.Timestamp(t).time()
+
     if df.empty:
         raise EmptyDataFrame()
+    # first filter based on the time-window
+    if time_window is not None:
+        t_start = parse_time(time_window[0])
+        t_end = parse_time(time_window[1])
+        if t_start > t_end:
+            # start before midnight
+            df = df[(df.index.time >= t_start) | (df.index.time < t_end)]
+        else:
+            df = df[(df.index.time >= t_start) & (df.index.time < t_end)]
+
     return df.resample(resolution).min()
 
 
+def share_of_standby(df, resolution='24h', time_window=None):
+    """
+    Compute the share of the standby power in the total consumption.
+
+    Parameters
+    ----------
+    df : Pandas DataFrame
+        Power (typically electricity, can be anything)
+    resolution : str, default='d'
+        Resolution of the computation.  Data will be resampled to this resolution (as mean) before computation
+        of the minimum.
+        String that can be parsed by the pandas resample function, example ='h', '15min', '6h'
+    time_window : tuple with start-hour and end-hour, default=None
+        Specify the start-time and end-time for the analysis.
+        Only data within this time window will be considered.
+        Both times have to be specified as string ('01:00', '06:30') or as datetime.time() objects
+
+    Returns
+    -------
+    fraction : float between 0-1 with the share of the standby consumption
+    """
+
+    df_ = pd.DataFrame(df)
+    p_sb = standby(df_, resolution, time_window)
+    df_resampled = df_.resample(resolution).mean()
+    p_tot = df_resampled.sum()
+    p_standby = p_sb.sum()
+    share_standby = p_standby/p_tot
+    return share_standby.iloc[0]
+
+
 def count_peaks(ts):
     """
     Toggle counter for gas boilers

opengrid/library/plotting.py

@@ -1,7 +1,10 @@
 import os
+import os
 import numpy as np
 import pandas as pd
 import matplotlib
+import pandas as pd
+import numpy as np
 import matplotlib.cm as cm
 import matplotlib.pyplot as plt
 from matplotlib.dates import date2num, num2date, HourLocator, DayLocator, AutoDateLocator, DateFormatter
@@ -19,6 +22,12 @@ def plot_style():
     matplotlib.style.use('seaborn-deep')
 
     plt.rcParams['figure.figsize'] = 16, 6
+
+    # To overrule the legend style
+    plt.rcParams['legend.facecolor'] = "#ffffff"
+    plt.rcParams['legend.frameon'] = True
+    plt.rcParams['legend.framealpha'] = 1
+
     return plt
 
 
@@ -114,3 +123,56 @@ def carpet(timeseries, **kwargs):
     plt.title(title)
 
     return im
+
+
+def boxplot(df, plot_mean=False, plot_ids=None, title=None, xlabel=None, ylabel=None):
+    """
+    Plot boxplots
+
+    Plot the boxplots of a dataframe in time
+
+    Parameters
+    ----------
+    df: Pandas Dataframe
+        Every collumn is a timeseries
+    plot_mean: bool
+        Wether or not to plot the means
+    plot_ids: [str]
+        List of id's to plot
+
+    Returns
+    -------
+    matplotlib figure
+    """
+
+    df = df.applymap(float)
+    description = df.apply(pd.DataFrame.describe, axis=1)
+
+    # plot
+    plt = plot_style()
+
+    plt.boxplot(df)
+    #plt.setp(bp['boxes'], color='black')
+    #plt.setp(bp['whiskers'], color='black')
+    if plot_ids is not None:
+        for id in plot_ids:
+            if id in df.columns:
+                plt.scatter(x=range(1, len(df) + 1), y=df[id], label=str(id))
+
+    if plot_mean:
+        plt.scatter(x=range(1, len(df) + 1), y=description['mean'], label="Mean", color='k', s=30, marker='+')
+
+    ax = plt.gca()
+    ax.set_xticklabels(df.index)
+    #plt.xticks(rotation=45)
+
+    plt.legend()
+    if title is not None:
+        plt.title(title)
+    if xlabel is not None:
+        plt.xlabel(xlabel)
+    if ylabel is not None:
+        plt.ylabel(ylabel)
+
+    return plt.gcf()
+

opengrid/tests/test_analyses.py

@@ -7,6 +7,7 @@
 
 import unittest
 import pandas as pd
+import numpy as np
 
 import opengrid as og
 from opengrid import datasets
@@ -22,6 +23,30 @@ def test_standby(self):
 
         self.assertRaises(EmptyDataFrame, og.analysis.standby, pd.DataFrame)
 
+    def test_standby_with_time_window(self):
+        df = datasets.get('elec_power_min_1sensor')
+        res = og.analysis.standby(df, 'D', time_window=('01:00', '06:00'))
+        self.assertEqual(res.index.tz.zone, 'Europe/Brussels')
+        self.assertEqual(res.to_json(), '{"1507327200000":61.739999936,"1507413600000":214.9799999222,"1507500000000":53.0399997951,"1507586400000":55.7399999164,"1507672800000":59.94000006,"1507759200000":69.4800002407,"1507845600000":56.8200000236,"1507932000000":54.1799997864,"1508018400000":54.779999801,"1508104800000":54.7199997772,"1508191200000":98.5199999576,"1508277600000":55.6799999066,"1508364000000":53.9399997052,"1508450400000":109.5599999931,"1508536800000":144.3600001093,"1508623200000":52.7999997279}')
+
+        res = og.analysis.standby(df, 'D', time_window=('22:00', '06:00'))
+        self.assertEqual(res.index.tz.zone, 'Europe/Brussels')
+        self.assertEqual(res.to_json(), '{"1507327200000":61.739999936,"1507413600000":119.2800000636,"1507500000000":53.0399997951,"1507586400000":55.7399999164,"1507672800000":59.94000006,"1507759200000":69.4800002407,"1507845600000":56.8200000236,"1507932000000":54.1799997864,"1508018400000":54.779999801,"1508104800000":54.7199997772,"1508191200000":98.5199999576,"1508277600000":55.6799999066,"1508364000000":53.9399997052,"1508450400000":96.3000000408,"1508536800000":133.9200000744,"1508623200000":52.7999997279}')
+
+    def test_share_of_standby_1(self):
+        df = pd.DataFrame(data={'conso':np.ones(48)},
+                          index=pd.DatetimeIndex(start=pd.Timestamp('20180304'), periods=48, freq='h'))
+        share_of_standby = og.analysis.share_of_standby(df, resolution='24h')
+        self.assertEqual(share_of_standby, 1.0)
+
+    def test_share_of_standby_2(self):
+        df = pd.DataFrame(data={'conso':np.ones(48)},
+                          index=pd.DatetimeIndex(start=pd.Timestamp('20180304'), periods=48, freq='h'))
+        df.iloc[0,0] = 0
+        share_of_standby = og.analysis.share_of_standby(df, resolution='24h')
+        self.assertAlmostEqual(share_of_standby, 0.5106382978723404)
+
+
     def test_count_peaks(self):
         df = datasets.get('gas_dec2016_min')
         ts = df['313b'].head(100)

opengrid/tests/test_plotting.py

@@ -18,13 +18,30 @@ def test_default(self):
 class CarpetTest(unittest.TestCase):
     def test_default(self):
         import numpy as np
-        index = pd.date_range('2015-1-1', '2015-12-31', freq='h')
+        index = pd.date_range('2015-1-1', '2015-2-1', freq='h')
         ser = pd.Series(np.random.normal(size=len(index)), index=index, name='abc')
         assert plotting.carpet(ser) is not None
 
     def test_empty(self):
         assert plotting.carpet(pd.Series(index=list('abc'))) is None
 
+class BoxplotTest(unittest.TestCase):
+    def test_default(self):
+        import numpy as np
+        import pandas as pd
+        from opengrid.library import plotting
+        index = pd.date_range('2015-1-1', '2015-2-1', freq='d')
+        df = pd.DataFrame(index=index, data=np.random.randint(5, size=(len(index),20)))
+        plotting.boxplot(df)
+
+    def test_arguments(self):
+        import numpy as np
+        import pandas as pd
+        from opengrid.library import plotting
+        index = pd.date_range('2015-1-1', '2015-2-1', freq='d')
+        df = pd.DataFrame(index=index, data=np.random.randint(5, size=(len(index),20)))
+        plotting.boxplot(df, plot_mean=True, plot_ids=[2, 3], title="Title", xlabel="xlable", ylabel="ylable")
+
 
 if __name__ == '__main__':
     unittest.main()