PCEA is the abbreviation for Python Compound Event Analysis.
Compound events are defined as the superposition of events within a specific time window, exhibiting six types of relationships as illustrated in the figure.
pyCEA is based on event analysis. It can identify events within signals based on specific thresholds and calculate the occurrence time, intensity, peak, and duration of those events.
It can also detect event chains (compound events) and calculate their probabilities. An event chain refers to two events that occur synchronously in time.
Examples include: El Niño—drought—wildfire chain events; heatwave—drought compound events; drought—flood compound events.
For the installation of pyECA, the following packages are required:
PCEA can be installed using pip
pip install PCEA
Two quick examples of pyCEA usage is as following.
- The data from data/demo.csv is read in, where the columns represent different variables and the rows represent sampling times.
import numpy as np
from PCEA import CEA
import pandas as pd
# Read the data in data/demo.csv, where the columns represent different
# variables and the rows represent sampling times.
ts = pd.read_csv("./data/demo.csv", index_col=0, header=0)
ts = ts.iloc[:, [0, 1]]
cea = CEA(ts, delta=6, threshold=[-np.inf, -0.5])
cea.run_cea(save_path='./data/resuls.xlsx') # run compound event analysis- Input a boolean matrix of size m x n.
from PCEA import CEA
import numpy as np
ts = np.random.choice([True, False], [720, 3]) # Generate a boolean matrix,
# where True represents the occurrence of an event.
cea = CEA(ts, delta=3, is_binary_array=True) # If the input is already a boolean matrix,
cea.run_cea(save_path='./data/resuls.xlsx') # run compound event analysis
# then you need to set "is_binary" to TRUE,
# and the threshold parameter "threshold" is not required.More examples can be seen in Jin et al (2025)