bee-health-calculation-model-bgood-wp2

Bee health calculation model B-GOOD WP2 This code was developed by Luke Chamberlain at Nottingham Trent University with the support of the B-GOOD project: https://b-good-project.eu/.

Description

Calculates the cumulative weight anomaly based on input weight data of two or more hives in JSON array format. Returns the first principal component score of each hive. A principal component analysis is done on the normalized cumulative sum of the daily mean weight of each hive.

Running the model

1. Install dependencies

• Python3: https://www.python.org/downloads/
• Numpy: https://numpy.org/
• Pandas: https://pypi.org/project/pandas/
• Scipy: https://scipy.org/

2. Run the model

python3

data=[
    {
    "time": "2023-11-16T00:00:00Z",
    "weight_1": 54.166666666666666,
    "weight_2": 52.125,
    "weight_3": 51.125
    },
    {
    "time": "2023-11-16T00:13:55Z",
    "weight_1": 46.166666666666666,
    "weight_3": 53.125
    },
    {
    "time": "2023-11-16T02:11:10Z",
    "weight_1": 43.446666666666666,
    "weight_2": 50.125
    }
]

cumulative_weight_anomaly(data)

Author

Luke Chamberlain [email protected]

Input data

Measurement

• net_weight_kg (weight excluding beekeeper actions)
• cumulative weight per day (set to 0 at midnight)

Data interval

• 15 min average

Data period

• relative interval of 7 * 24 hours

Data items

• all hives of the apiary as a single data stream

Output data

• PCA of data (https://github.com/laxip/PCAphp)
• deviation of comparison of each hive feature