In [49]:

    

import folium
import pandas as pd
import numpy as np
import os
from geopy.geocoders import Nominatim
import geocoder
import matplotlib.pyplot as plt
import json
from folium import plugins
import branca.colormap as cm
from tqdm import tqdm_notebook
plt.style.use('ggplot')


    

In [2]:

    

def get_elevation(lat,lon):
    try:
        return geocoder.google([lat, lon], method='elevation', key=os.environ['GOOGLE_API_KEY']).meters
    except:
        return np.NaN


    

In [3]:

    

def read_csv(filename):
    df = pd.read_csv(filename)
    NewLap_idx = df[df.Time == 'New Lap'].index.values
    df['Time'] = pd.to_datetime(df['Time'], errors='coerce')
    df.dropna(how='any',axis=0, subset=['Time'], inplace=True)
    for lab in ['LongitudeDegrees', 'LatitudeDegrees', 'AltitudeMeters', 'heartratebpm/value']:
        df[lab] = df[lab].interpolate()
    df['dist'] = \
        haversine_np(df.LongitudeDegrees.shift(), df.LatitudeDegrees.shift(),
                     df.loc[1:, 'LongitudeDegrees'], df.loc[1:, 'LatitudeDegrees'])
    df['cum_dis'] = df.dist.cumsum()
    df['DeltaTime'] = (df.Time - df.Time.shift()).dt.total_seconds()
    df['Speed'] = df.apply(lambda x: x['dist']/float(x['DeltaTime'])*3600.0 if x['DeltaTime'] > 0.5 else np.NaN, axis=1)
    df['Speed'] = df.apply(lambda x: x['Speed'] if x['Speed'] < 20 else np.NaN, axis=1)
    df['DeltaAltitude'] = df.AltitudeMeters - df.AltitudeMeters.shift()
    df.to_csv('output/csv/' + os.path.basename(filename))
    return df, NewLap_idx


    

In [4]:

    

def haversine_np(lon1, lat1, lon2, lat2):
    """
    Calculate the great circle distance between two points
    on the earth (specified in decimal degrees)

    All args must be of equal length.    

    """
    lon1, lat1, lon2, lat2 = map(np.radians, [lon1, lat1, lon2, lat2])

    dlon = lon2 - lon1
    dlat = lat2 - lat1

    a = np.sin(dlat/2.0)**2 + np.cos(lat1) * np.cos(lat2) * np.sin(dlon/2.0)**2

    c = 2 * np.arcsin(np.sqrt(a))
    km = 6367 * c
    return km


    

In [4]:

    

#path = 'data/csv/'
path = 'output/csv'
geolocator = Nominatim()
allFiles = os.listdir(path)
frame = pd.DataFrame(columns=('Date', 'Filename', 'Kilometers', 'Speed', 'Time', 'Climb', 'Lat', 'Long', 'Adress'))
dict_df = dict()
#dict_newlap = dict()
for i, file in tqdm_notebook(enumerate(allFiles)):
    df = pd.read_csv('output/csv/' + file, index_col=0, parse_dates=['Time'])
    #df, newlap = read_csv(path + file)
    dict_df[df['Time'][0].strftime("%Y-%m-%d")] = df
    location = geolocator.reverse(str(df.LatitudeDegrees.mean()) + "," + str(df.LongitudeDegrees.mean()))[0]
    frame.loc[i] = [df['Time'][0], file, df.cum_dis.max(), df.Speed.mean(), df.DeltaTime.sum()/60,
                    df['DeltaAltitude'][df['DeltaAltitude'] > 0].sum(),
                    df.LatitudeDegrees.mean(), df.LongitudeDegrees.mean(), location]


    

In [5]:

    

frame.loc[frame[frame['Filename'] == '2017-09-01T05_07_50Z_PT50M7S_Running.csv'].index,'Speed'] = 10.9
frame.loc[frame[frame['Filename'] == '2017-09-01T05_07_50Z_PT50M7S_Running.csv'].index,'Kilometers'] = 8.09
frame.loc[frame[frame['Filename'] == '2017-09-01T05_07_50Z_PT50M7S_Running.csv'].index,'Time'] = 44.5
frame.loc[frame[frame['Filename'] == '2017-09-01T05_07_50Z_PT50M7S_Running.csv'].index,'Climb'] = 130


    

In [6]:

    

frame.sort_values('Date', ascending=False)


    

    
Out[6]:







  

    

      

      
Date
      
Filename
      
Kilometers
      
Speed
      
Time
      
Climb
      
Lat
      
Long
      
Adress
    
  
  

    

      
31
      
2017-09-28 17:31:41.836
      
2017-09-28T17_32_18Z_PT40M33S_Running.csv
      
7.643801
      
11.602275
      
41.166867
      
742.000
      
48.822395
      
2.338011
      
Le Petit Train, Allée du Bardo, Montsouris, Pa...
    
    

      
1
      
2017-09-26 17:29:38.784
      
2017-09-26T17_29_38Z_PT42M55S_Running.csv
      
8.761183
      
11.979408
      
42.932483
      
890.000
      
48.822418
      
2.337851
      
Le Petit Train, Allée du Bardo, Montsouris, Pa...
    
    

      
53
      
2017-09-23 09:23:27.402
      
2017-09-23T09_23_58Z_PT1H59M11S_Running.csv
      
21.848359
      
11.175643
      
119.718800
      
2166.000
      
48.780568
      
2.292178
      
Avenue Charles Péguy, Sceaux, Antony, Hauts-de...
    
    

      
45
      
2017-09-21 16:34:23.641
      
2017-09-21T16_34_23Z_PT45M38S_Running.csv
      
10.592581
      
11.908545
      
45.635200
      
1048.000
      
48.846442
      
2.335671
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
41
      
2017-09-19 16:36:08.981
      
2017-09-19T16_36_08Z_PT44M32S_Running.csv
      
9.041560
      
12.643044
      
44.544367
      
897.000
      
48.846427
      
2.335698
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
5
      
2017-09-16 07:17:00.070
      
2017-09-16T07_17_00Z_PT1H30M8S_Running.csv
      
17.096634
      
11.390383
      
90.146483
      
1854.000
      
48.784790
      
2.291165
      
Rue François Moreau, Fontenay-aux-Roses, Anton...
    
    

      
6
      
2017-09-14 17:23:15.896
      
2017-09-14T16_20_43Z_PT1H42M47S_Biking,Running...
      
8.048995
      
11.910404
      
36.919733
      
848.000
      
48.822385
      
2.338098
      
Le Petit Train, Allée du Bardo, Montsouris, Pa...
    
    

      
37
      
2017-09-12 16:59:14.957
      
2017-09-12T16_59_14Z_PT51M33S_Running.csv
      
9.248593
      
12.426092
      
51.559717
      
909.000
      
48.846306
      
2.335477
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
50
      
2017-09-09 08:27:03.960
      
2017-09-09T08_27_03Z_PT1H55M7S_Running.csv
      
21.728122
      
11.392283
      
115.132050
      
2317.000
      
48.779607
      
2.291531
      
Fontenay-Houdan, Rue de Fontenay, Sceaux, Anto...
    
    

      
24
      
2017-09-07 17:01:55.000
      
RunnerUp_0058.csv
      
8.679763
      
12.729851
      
41.800000
      
853.000
      
48.846475
      
2.335826
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
47
      
2017-09-05 17:06:22.000
      
RunnerUp_0057.csv
      
8.958269
      
12.844046
      
41.433333
      
822.000
      
48.846517
      
2.335812
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
21
      
2017-09-04 17:38:54.000
      
RunnerUp_0056.csv
      
3.398560
      
11.900422
      
17.116667
      
301.000
      
48.844327
      
2.335079
      
15, Rue Auguste Comte, Odéon, 6e, Paris, Île-d...
    
    

      
38
      
2017-09-02 08:52:00.450
      
2017-09-02T08_52_00Z_PT1H58M34S_Running.csv
      
21.516255
      
8.859990
      
121.576833
      
1717.000
      
48.777818
      
2.290877
      
EHPAD Résidence Renaudin, Passage Marguerite R...
    
    

      
27
      
2017-09-01 05:07:50.590
      
2017-09-01T05_07_50Z_PT50M7S_Running.csv
      
8.090000
      
10.900000
      
44.500000
      
130.000
      
48.842785
      
2.329997
      
Starbucks Coffee, Rue Jules Chaplain, ND-des-C...
    
    

      
15
      
2017-08-30 17:20:06.174
      
Cristian_Perez_2017-08-30_19-20-05.csv
      
8.146645
      
11.694941
      
41.716667
      
121.772
      
48.846472
      
2.335846
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
43
      
2017-08-29 17:15:33.209
      
Cristian_Perez_2017-08-29_19-15-32.csv
      
8.462380
      
11.721614
      
42.616667
      
164.441
      
48.846504
      
2.335784
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
32
      
2017-08-26 08:07:54.000
      
RunnerUp_0051.csv
      
23.109283
      
10.833283
      
122.233333
      
2713.000
      
48.778201
      
2.291200
      
EHPAD Résidence Renaudin, Passage Marguerite R...
    
    

      
54
      
2017-08-23 17:28:48.000
      
RunnerUp_0050.csv
      
5.046360
      
12.262504
      
24.100000
      
515.000
      
48.846125
      
2.335200
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
28
      
2017-08-22 17:48:37.000
      
RunnerUp_0049.csv
      
7.042143
      
11.952149
      
33.416667
      
825.000
      
48.846353
      
2.335627
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
39
      
2017-08-21 17:50:10.000
      
RunnerUp_0048.csv
      
7.073655
      
12.216654
      
32.850000
      
801.000
      
48.846266
      
2.335489
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
11
      
2017-08-19 08:43:04.000
      
RunnerUp_0047.csv
      
17.425455
      
11.227346
      
89.066667
      
1879.000
      
48.785283
      
2.290892
      
Rue François Moreau, Fontenay-aux-Roses, Anton...
    
    

      
22
      
2017-08-17 18:03:58.000
      
RunnerUp_0046.csv
      
5.329528
      
12.031507
      
25.083333
      
581.000
      
48.846002
      
2.335284
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
25
      
2017-08-16 17:46:09.000
      
RunnerUp_0044.csv
      
7.007194
      
12.314706
      
32.766667
      
774.000
      
48.846335
      
2.335589
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
55
      
2017-08-14 10:08:22.000
      
RunnerUp_43.csv
      
6.288584
      
11.746766
      
30.150000
      
654.000
      
48.822337
      
2.338085
      
Le Petit Train, Allée du Bardo, Montsouris, Pa...
    
    

      
4
      
2017-08-12 08:00:36.000
      
RunnerUp_0042.csv
      
20.083537
      
10.992925
      
97.766667
      
2349.000
      
48.782023
      
2.290309
      
16, Boulevard Desgranges, Sceaux, Antony, Haut...
    
    

      
13
      
2017-08-10 18:25:48.000
      
RunnerUp_0040.csv
      
5.048189
      
11.765151
      
25.500000
      
470.000
      
48.846094
      
2.335153
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
0
      
2017-08-08 18:01:53.000
      
RunnerUp_0038.csv
      
5.364445
      
11.557867
      
26.633333
      
609.000
      
48.846072
      
2.335317
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
34
      
2017-08-06 10:10:35.000
      
RunnerUp_0037.csv
      
7.722241
      
11.964914
      
36.833333
      
801.000
      
48.846405
      
2.335812
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
35
      
2017-08-03 17:59:08.000
      
RunnerUp_0036.csv
      
5.534886
      
11.573736
      
27.000000
      
639.000
      
48.846058
      
2.335346
      
Poste des surveillants du jardin, Place André ...
    
    

      
14
      
2017-07-30 09:42:55.000
      
RunnerUp_0035.csv
      
7.526653
      
11.697312
      
36.283333
      
894.000
      
48.846395
      
2.335657
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
30
      
2017-07-27 18:20:01.000
      
RunnerUp_0034.csv
      
5.841380
      
11.599974
      
29.850000
      
628.000
      
48.845605
      
2.335139
      
Comtesse de Ségur, Rue Auguste Comte, Odéon, 6...
    
    

      
46
      
2017-07-25 19:50:41.000
      
RunnerUp_0033.csv
      
5.027124
      
11.628791
      
22.233333
      
633.000
      
48.846215
      
2.335679
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
23
      
2017-07-24 20:54:20.000
      
RunnerUp_0032.csv
      
4.530297
      
11.905375
      
21.183333
      
443.000
      
48.846382
      
2.335684
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
12
      
2017-07-19 19:55:02.000
      
RunnerUp_31.csv
      
4.006633
      
11.753229
      
19.050000
      
537.000
      
48.849479
      
2.332601
      
Sœur, Rue Honoré Chevalier, Odéon, 6e, Paris, ...
    
    

      
48
      
2017-07-18 20:14:04.000
      
RunnerUp_0030.csv
      
4.014472
      
11.086862
      
26.450000
      
486.000
      
48.844326
      
2.334989
      
15, Rue Auguste Comte, Odéon, 6e, Paris, Île-d...
    
    

      
42
      
2017-07-11 18:09:12.000
      
RunnerUp_0029.csv
      
5.035661
      
11.444715
      
25.900000
      
495.000
      
48.846123
      
2.335252
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
2
      
2017-07-10 18:30:43.000
      
RunnerUp_0028.csv
      
5.086652
      
11.832552
      
24.866667
      
528.000
      
48.846127
      
2.335270
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
16
      
2017-07-09 09:53:50.000
      
RunnerUp_0027.csv
      
8.109093
      
11.397367
      
37.083333
      
817.000
      
48.846400
      
2.335600
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
52
      
2017-07-06 18:12:50.000
      
RunnerUp_0026.csv
      
4.665151
      
9.431150
      
30.233333
      
631.000
      
48.843965
      
2.333494
      
57b, Rue d'Assas, Odéon, 6e, Paris, Île-de-Fra...
    
    

      
29
      
2017-07-05 18:35:48.000
      
RunnerUp_0025.csv
      
5.030024
      
11.446046
      
24.616667
      
521.000
      
48.846461
      
2.335721
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
17
      
2017-07-04 17:42:55.000
      
RunnerUp_24.csv
      
5.209022
      
11.345380
      
27.383333
      
513.000
      
48.846047
      
2.335238
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
7
      
2017-06-18 09:39:15.000
      
RunnerUp_0023.csv
      
5.019963
      
10.748088
      
27.833333
      
547.000
      
48.846146
      
2.335195
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
49
      
2017-05-28 10:27:13.000
      
RunnerUp_22.csv
      
5.021169
      
11.038293
      
26.450000
      
571.000
      
48.846203
      
2.335309
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
51
      
2017-05-21 09:57:30.000
      
RunnerUp_0021.csv
      
5.031971
      
11.359628
      
25.750000
      
527.000
      
48.846216
      
2.335359
      
Blanche de Castille, Cour de la Présidence, Od...
    
    

      
10
      
2017-04-30 09:52:48.000
      
RunnerUp_0020.csv
      
5.017558
      
11.314596
      
25.783333
      
554.000
      
48.846171
      
2.335251
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
44
      
2017-04-17 11:01:08.000
      
RunnerUp_0019.csv
      
5.012422
      
11.671251
      
24.950000
      
510.000
      
48.846221
      
2.335243
      
Théâtre du Luxembourg, Rue Auguste Comte, Odéo...
    
    

      
8
      
2017-04-16 10:38:27.000
      
RunnerUp_0018.csv
      
4.602083
      
11.295570
      
23.600000
      
470.000
      
48.846213
      
2.335456
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
40
      
2017-03-26 11:59:32.000
      
RunnerUp_0017.csv
      
5.688288
      
11.581280
      
34.716667
      
577.000
      
48.845600
      
2.335287
      
Pierre Mendès France, Rue Auguste Comte, Odéon...
    
    

      
20
      
2017-03-25 10:01:18.000
      
RunnerUp_0016.csv
      
4.629997
      
11.634636
      
23.100000
      
459.000
      
48.846481
      
2.335620
      
Blanche de Castille, Cour de la Présidence, Od...
    
    

      
26
      
2016-10-08 05:20:17.000
      
RunnerUp_10.csv
      
15.028409
      
8.991190
      
100.200000
      
1486.000
      
36.450690
      
25.415905
      
Θήρα - Οία, Φοινικιά, Δήμος Θήρας, Περιφερειακ...
    
    

      
18
      
2016-10-04 09:45:14.000
      
RunnerUp_0007.csv
      
0.005205
      
6.246022
      
0.050000
      
4.000
      
48.725136
      
2.145281
      
458, Avenue des Accélérateurs, CEA Saclay, Sai...
    
    

      
36
      
2016-10-02 09:31:51.000
      
RunnerUp_0006.csv
      
7.001115
      
11.774183
      
34.916667
      
655.000
      
48.846376
      
2.335594
      
Anne de Beaujeu, Cour de la Présidence, Odéon,...
    
    

      
9
      
2016-10-01 08:02:18.000
      
RunnerUp_0005.csv
      
4.008485
      
11.110023
      
20.733333
      
481.000
      
48.846831
      
2.336178
      
Jardin du Luxembourg, Cour de Jonction, Odéon,...
    
    

      
19
      
2016-09-13 19:59:11.000
      
RunnerUp_0004.csv
      
2.004824
      
10.999175
      
10.866667
      
153.000
      
43.601009
      
3.897712
      
Banque Populaire, Rue de Syracuse, Port Marian...
    
    

      
33
      
2016-09-13 19:49:33.000
      
RunnerUp_0002.csv
      
1.037511
      
11.742230
      
5.150000
      
111.000
      
43.607851
      
3.889808
      
Hortus, 38, Place du Millénaire, Antigone, Cen...
    
    

      
3
      
2016-09-10 06:16:10.000
      
RunnerUp_0001.csv
      
4.040153
      
10.275758
      
23.133333
      
409.000
      
48.845935
      
2.335594
      
Poste des surveillants du jardin, Place André ...
    
  

In [7]:

    

recent_frame = frame[frame['Date'] > pd.datetime(2017, 6, 1)]


    

In [81]:

    

recent_frame = recent_frame.assign(Month=recent_frame['Date'].dt.month, Week=recent_frame['Date'].dt.week, 
                                   Weekday=recent_frame['Date'].dt.weekday)


    

In [78]:

    

months = ['June', 'July', 'August', 'Sept']
fig, axr = plt.subplots(1,3, figsize=(14,5))
axr[0].bar(recent_frame.groupby(recent_frame.Date.dt.month).agg({'Kilometers': 'sum'}).index, 
        recent_frame.groupby(recent_frame.Date.dt.month).agg({'Kilometers': 'sum'}).values,
        yerr=recent_frame.groupby(recent_frame.Date.dt.month).agg({'Kilometers': 'std'}).values,
        tick_label=months, label='Number of kilometers')
axr[0].set_title('Total number of kilometers')
axr[0].set_ylabel('Km')
recent_frame.boxplot(['Speed'], by='Month', ax=axr[1])
axr[1].set_xticklabels(months)
axr[1].set_ylabel('Km/h')
recent_frame.groupby('Month')['Speed'].count().plot.bar(ax=axr[2], color='C1')
axr[2].set_xticklabels(months)
axr[2].set_title('Number of trainings')
plt.show()


    

In [93]:

    

months = ['June', 'July', 'August', 'Sept']
fig, axr = plt.subplots(3, figsize=(14,8), sharex=True)
recent_frame.groupby('Week')['Kilometers'].sum().plot.bar(ax=axr[0])
axr[0].set_title('Total number of kilometers')
axr[0].set_ylabel('Km')
recent_frame.boxplot(['Speed'], by='Week', ax=axr[1])
axr[1].set_ylabel('Km/h')
recent_frame.groupby('Week')['Speed'].count().plot.bar(ax=axr[2], color='C1')
axr[2].set_title('Number of trainings')
plt.show()


    

In [118]:

    

weekdays = [ 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
fig, axr = plt.subplots(3, figsize=(14,8))
recent_frame.groupby('Weekday')['Kilometers'].sum().plot.area(ax=axr[0])
axr[0].set_title('Total number of kilometers')
axr[0].set_ylabel('Km')
axr[0].set_xticklabels(weekdays)
recent_frame.boxplot(['Speed'], by='Weekday', ax=axr[1])
axr[1].set_ylabel('Km/h')
axr[1].set_xticklabels(weekdays)
recent_frame.groupby('Weekday')['Speed'].count().plot.area(ax=axr[2], color='C1')
axr[2].set_title('Number of trainings')
axr[2].set_xticklabels(weekdays)
plt.show()


    

In [119]:

    

from sklearn import datasets, linear_model
from sklearn.metrics import mean_squared_error, r2_score


    

In [120]:

    

recent_frame[['Kilometers', 'Climb']].values.shape


    

    
Out[120]:





(42, 2)

In [121]:

    

np.stack([recent_frame.index.values,recent_frame['Time'].values, recent_frame['Climb'].values], axis=1).shape


    

    
Out[121]:





(42, 3)

In [122]:

    

#X_train = np.stack([recent_frame.index.values, recent_frame['Kilometers'].values, recent_frame['Climb'].values ], axis=1)
X_train = recent_frame[['Kilometers','Climb']].values
y_train = recent_frame['Time']
X_test = np.array([42.195, 400]).reshape(1,-1)


    

In [123]:

    

regr = linear_model.LinearRegression()
regr.fit(X_train, y_train)
y_pred = regr.predict(X_test)
print('Distance: {} Km, Speed: {} Km/h, Time: {} hours \n'.format(X_test[0][0], X_test[0][0]/float(y_pred[0]/60), y_pred[0]/60))


    

    




Distance: 42.195 Km, Speed: 10.563563454086022 Km/h, Time: 3.99439073598586 hours 

In [124]:

    

Dis = 10
Climb = 400
val = np.array([Dis, Climb]).reshape(1,-1)
print(Dis/(regr.predict(val)/60))
#print('Distance: {} Km, Speed: {} Km/h, Time: {} hours \n'.format(X_test[0][1], X_test[0][1]/float(y_pred[0]/60), y_pred[0]/60))


    

    




[ 11.210528]

In [221]:

    

plt.figure(figsize=(15,4))
plt.scatter(X_train[:,0], regr.predict(X_train)/60)
plt.scatter(X_test[:,0], y_pred/60)
plt.plot([0, X_test[:,0]], [0, y_pred/60], '--', color='C2')
plt.ylabel('Hours')
plt.xlabel('Kilometers')
plt.show()


    

In [223]:

    

plt.figure(figsize=(15,4))
ax = plt.axes()
ax.set_xlim([pd.datetime(2017,6,15),pd.datetime(2017,10,10)])

legend_sizes = np.sort(recent_frame.Climb)[::len(recent_frame.Climb)//4][-3:]
# get the indices for each of the legend sizes
indices = [np.where(recent_frame.Climb==v)[0][0] for v in legend_sizes]
# plot each point again, and its value as a label
for i in indices:
    plt.scatter(x=recent_frame.Date.iloc[i], y=recent_frame.Speed.iloc[i], c=recent_frame.Kilometers.iloc[i], s=recent_frame.Climb.iloc[i]/2,
                  vmin=0,vmax=1,edgecolors='none',
                  label='{:.2f}'.format(recent_frame.Climb.iloc[i]))
# add the legend
plt.scatter(x=recent_frame.Date.values, y=recent_frame.Speed, c=recent_frame.Kilometers, s=recent_frame.Climb, label=None)
#plt.scatter(x=pd.to_datetime('2017-10-09', format='%Y-%m-%d'), y=X_test[0][0]/float(y_pred[0]/60), c=X_test[0][0], s=X_test[0][1], linewidth=3)
plt.plot(X_test[0][0]/float(y_pred[0]/60))
plt.colorbar(label='Distance')
plt.xlabel('Date')
plt.ylabel('Average Speed')
plt.legend(scatterpoints=1, title='Climb', labelspacing=1.2)
plt.show()


    

In [261]:

    

list_speed_bubble = list()
list_speed_prediction = list()
list_linear_regression = list()
for k,val in recent_frame.iterrows():
    list_speed_bubble.append({'x': val['Date'].strftime("%Y-%m-%d"), 'y': np.round(val['Speed'],2), 'r': np.round(val['Kilometers'],2)})
    to_predict = val[['Kilometers', 'Climb']].values.reshape(1,-1)
    list_linear_regression.append({'x': np.round(val['Kilometers'],2), 'y': np.round(regr.predict(to_predict)[0]/60,2)})
    #list_speed_prediction.append({'x': val['Date'].strftime("%Y-%m-%d"), 'y': np.round(val['Kilometers']/(regr.predict(to_predict)[0]/60),2), 'r': np.round(val['Kilometers'],2)})

list_speed_bubble.append({'x': '2017-10-08', 'y': np.round(X_test[0][0]/float(y_pred[0]/60),2), 'r': np.round(X_test[0][0],2)})
list_speed_prediction.append({'x': '2017-10-08', 'y': np.round(X_test[0][0]/float(y_pred[0]/60),2), 'r': np.round(X_test[0][0],2)})


    

In [262]:

    

to_predict = np.reshape([0,40], (1,-1))
regr.predict(to_predict)[0]/60
#list_speed_predi


    

    
Out[262]:





-0.048497971407600253

In [263]:

    

df_steps = pd.concat([recent_frame.Date.dt.strftime("%s"), recent_frame.Kilometers.round(1)], axis=1)

dic_steps = dict()
for k,x in df_steps.iterrows():
    dic_steps[x['Date']] = x['Kilometers']


    

In [264]:

    

with open('output/marathon.json', 'w') as outfile:
    json.dump({'weekday': {'Kilometers': [i[0] for i in recent_frame.groupby(recent_frame.Date.dt.weekday).agg({'Kilometers': 'mean'}).round(1).values.tolist()],
                           'Speed': [i[0] for i in recent_frame.groupby(recent_frame.Date.dt.weekday).agg({'Speed': 'mean'}).round(1).values.tolist()],
                           'Trains': [i[0] for i in recent_frame.groupby(recent_frame.Date.dt.weekday).agg({'Kilometers': 'count'}).round(1).values.tolist()],
                           'index' : weekdays
                          },
               'week': {'index' : [float(i) for i in recent_frame.groupby(recent_frame.Date.dt.week).agg({'Kilometers': 'count'}).index.tolist()],
                       'Kilometers': [float(i[0]) for i in recent_frame.groupby(recent_frame.Date.dt.week).agg({'Kilometers': 'sum'}).round(1).values],
                       'Speed': [float(i[0]) for i in recent_frame.groupby(recent_frame.Date.dt.week).agg({'Speed': 'mean'}).round(1).values],
                       'Trains': [float(i[0]) for i in recent_frame.groupby(recent_frame.Date.dt.week).agg({'Kilometers': 'count'}).round(1).values]
                       },
               'month': {'Kilometers' : [float(i) for i in recent_frame.groupby(recent_frame.Date.dt.month).agg({'Kilometers': 'sum'}).round(1).values],
                         'Speed' : [float(i) for i in recent_frame.groupby(recent_frame.Date.dt.month).agg({'Speed': 'mean'}).round(1).values],
                         'Trains' : [float(i) for i in recent_frame.groupby(recent_frame.Date.dt.month).agg({'Kilometers': 'count'}).round(1).values],
                         'index' : months
               },
               'heatmap_km': dic_steps,
               'bubble_speed': {'real': list_speed_bubble, 'predict': sorted(list_speed_prediction, key=lambda x : x['x'], reverse=False)},
               'linear_reg': {'points': list_linear_regression, 'line': [{'x': 0, 'y': -0.04}, {'x':  X_test[:,0][0], 'y' : np.round((y_pred/60)[0], 2)}]}
               #'leaflet': [[i[0], i[1], 1] for i in [[float(v.Lat), float(v.Long)] for k,v in recent_frame.iterrows()]],
              }, outfile, indent=1)


    

In [254]:

    

mean_lat = np.mean([i[0] for i in [[float(v.Lat), float(v.Long)] for k,v in recent_frame.iterrows()]])
mean_long = np.mean([i[1] for i in [[float(v.Lat), float(v.Long)] for k,v in recent_frame.iterrows()]])
m = folium.Map(location=[mean_lat, mean_long ], zoom_start=11, tiles='OpenStreetMap')
m.add_child(plugins.HeatMap(recent_frame[['Lat','Long']].values, radius = 20), name='Training location')
colormap = cm.LinearColormap(['blue', 'green',  'yellow', 'orange', 'red'],  index=[0, 0.25, 0.5, 0.75, 1.0]).scale(0,20)
colormap.caption = 'Number of trainings'
m.add_child(colormap)
m


    

    
Out[254]:

In [ ]:

    




    

In [ ]: