Analyzing data from Movebank

[1]:

import movekit as mkit

Read data input

[2]:

data = mkit.read_movebank('datasets/Mangabey_monkeys_in_Uganda.csv')
data.head()

[2]:
time animal_id x y event-id visible location-long location-lat behavioural-classification comments study-specific-measurement sensor-type individual-taxon-canonical-name tag-local-identifier study-name
1337 1972-05-01 05:15:00 BT 206205.00004 61229.999999 1215038324 True 30.360556 0.553375 ? NaN SUBADULT natural-mark Lophocebus albigena BT Mangabey monkeys in Uganda (data from Waser 1975)
0 1972-05-01 05:15:00 KT 206220.00004 61219.999999 1215038318 True 30.360691 0.553285 ? NaN ADULT natural-mark Lophocebus albigena KT Mangabey monkeys in Uganda (data from Waser 1975)
710 1972-05-01 05:15:00 LN 206210.00004 61234.999999 1215038323 True 30.360601 0.553420 ? NaN ADULT natural-mark Lophocebus albigena LN Mangabey monkeys in Uganda (data from Waser 1975)
9637 1972-05-01 05:15:00 M Group 206209.00004 61220.499999 1099185632 True 30.360592 0.553289 NaN 9.0 NaN natural-mark Lophocebus albigena M Group Mangabey monkeys in Uganda (data from Waser 1975)
2148 1972-05-01 05:45:00 HK 206235.00004 61574.999999 1215038325 True 30.360824 0.556493 FEED:DS:FRUIT NaN ADULT natural-mark Lophocebus albigena HK Mangabey monkeys in Uganda (data from Waser 1975)

Basic Data Analysis

For further data analysis functions of movekit please refer to the other example notebooks or the documentation. It is important to preprocess the data from Movebank, as the data often contains duplicates which can lead to problems in further analysis steps.

[10]:

data = mkit.preprocess(data, interpolation = True, date_format = True)  # important to indicate the date format of time by setting date_format=True when applying interpolation
data.head()

Total number of missing values =  14437
comments                           9637
behavioural-classification         2400
study-specific-measurement         2400
time                                  0
animal_id                             0
x                                     0
y                                     0
event-id                              0
visible                               0
location-long                         0
location-lat                          0
sensor-type                           0
individual-taxon-canonical-name       0
tag-local-identifier                  0
study-name                            0
dtype: int64

[10]:
time animal_id x y event-id visible location-long location-lat behavioural-classification comments study-specific-measurement sensor-type individual-taxon-canonical-name tag-local-identifier study-name
1337 1972-05-01 05:15:00 BT 206205.00004 61229.999999 1215038324 True 30.360556 0.553375 ? NaN SUBADULT natural-mark Lophocebus albigena BT Mangabey monkeys in Uganda (data from Waser 1975)
0 1972-05-01 05:15:00 KT 206220.00004 61219.999999 1215038318 True 30.360691 0.553285 ? NaN ADULT natural-mark Lophocebus albigena KT Mangabey monkeys in Uganda (data from Waser 1975)
710 1972-05-01 05:15:00 LN 206210.00004 61234.999999 1215038323 True 30.360601 0.553420 ? NaN ADULT natural-mark Lophocebus albigena LN Mangabey monkeys in Uganda (data from Waser 1975)
9637 1972-05-01 05:15:00 M Group 206209.00004 61220.499999 1099185632 True 30.360592 0.553289 NaN 9.0 NaN natural-mark Lophocebus albigena M Group Mangabey monkeys in Uganda (data from Waser 1975)
2148 1972-05-01 05:45:00 HK 206235.00004 61574.999999 1215038325 True 30.360824 0.556493 FEED:DS:FRUIT 10.0 ADULT natural-mark Lophocebus albigena HK Mangabey monkeys in Uganda (data from Waser 1975) 
[13]:

data = mkit.extract_features(data)
data.head()

Extracting all absolute features: 100%|██████████| 100.0/100 [00:03<00:00, 32.24it/s]

[13]:
time animal_id x y distance direction turning average_speed average_acceleration stopped ... visible location-long location-lat behavioural-classification comments study-specific-measurement sensor-type individual-taxon-canonical-name tag-local-identifier study-name
0 1972-05-01 05:15:00 BT 206205.00004 61229.999999 0.0 (0.0, 0.0) 0.0 144.909354 -6.750061 0 ... True 30.360556 0.553375 ? 0.0 SUBADULT natural-mark Lophocebus albigena BT Mangabey monkeys in Uganda (data from Waser 1975)
1 1972-05-01 05:15:00 KT 206220.00004 61219.999999 0.0 (0.0, 0.0) 0.0 89.757240 8.491764 0 ... True 30.360691 0.553285 ? 0.0 ADULT natural-mark Lophocebus albigena KT Mangabey monkeys in Uganda (data from Waser 1975)
2 1972-05-01 05:15:00 LN 206210.00004 61234.999999 0.0 (0.0, 0.0) 0.0 280.498911 3.702460 0 ... True 30.360601 0.553420 ? 0.0 ADULT natural-mark Lophocebus albigena LN Mangabey monkeys in Uganda (data from Waser 1975)
3 1972-05-01 05:15:00 M Group 206209.00004 61220.499999 0.0 (0.0, 0.0) 0.0 41.606821 -0.149840 0 ... True 30.360592 0.553289 0 9.0 0 natural-mark Lophocebus albigena M Group Mangabey monkeys in Uganda (data from Waser 1975)
4 1972-05-01 05:45:00 HK 206235.00004 61574.999999 0.0 (0.0, 0.0) 0.0 243.150166 -7.612515 0 ... True 30.360824 0.556493 FEED:DS:FRUIT 10.0 ADULT natural-mark Lophocebus albigena HK Mangabey monkeys in Uganda (data from Waser 1975)

5 rows × 21 columns

[7]:

#calculate for each animal the distance covered between 1972-5-1 and 1972-10-1 12:00:00
mkit.distance_by_time(data, '1972-5-1','1972-10-1 12:00:00')

[7]:
animal_id distance
0 BB 29292.196235
1 BC 32339.850273
2 BR 25428.128415
3 BT 56379.253280
4 CH 0.000000
5 FB 0.000000
6 HK 44779.676823
7 JA 0.000000
8 JE 42121.957235
9 KT 44881.459185
10 LN 36373.411270
11 M Group 69659.023543
12 O 48254.273757
13 PB 16140.353833
14 RF 55879.886570
15 S 17618.932566
16 SK 22967.788494
17 TK 49030.622798

Create interactive maps with the geodata

Using the geodata interactive maps can be created. By clicking on the marker points the animal_id and the time of each marker is displayed.

[3]:

# as the creation of the map is very time-confusing it is recommended to select a subsample of animals for which the map is created.
mkit.plot_geodata(data, animal_list=["BT", "KT"])

Plotting geo data for specified animals: 100%|██████████| 2/2 [00:00<00:00,  4.34it/s]

[3]:
Make this Notebook Trusted to load map: File -> Trust Notebook

Additionally the movement of individual animals can be retraced by including movement lines in between the different markers.

[4]:

# each movement line has a tooltip to follow the sequence of movement (here animal "JA" for example starts in south west)
movement_map = mkit.plot_geodata(data, animal_list=["JA"], movement_lines=True)
movement_map

Plotting geo data for specified animals: 100%|██████████| 1/1 [00:00<00:00,  8.70it/s]

[4]:
Make this Notebook Trusted to load map: File -> Trust Notebook

The created maps can also be saved locally.

[10]:

#mkit.save_geodata_map(movement_map, "mymap.html")