I am struggling to compile and use X-13ARIMA-SEAT on a Mac with an m3 chip, and called it using sm.tsa.x13_arima_analysis(s, x12path='/usr/local/bin/x13as') in statsmodels. Finally, I succeeded and now share the method step by step.
In addition, I would like to know how to use Python libraries such as statsmodels or other seasonal adjustment tools using tramo-seats to seasonally adjust data?
Firstly we can download the X-13ARIMA-SEATS binary from the link SourceForge for macOS Apple Silicon (ARM64) and install it. Here are the detailed steps:
Step 1: Download X-13ARIMA-SEATS for macOS Apple Silicon
Download the X-13ARIMA-SEATS binary for macOS Apple Silicon from SourceForge:
curl -L -o x13as-mac-arm64.tar.gz https://sourceforge.net/projects/gretl/files/x13as/x13as-mac-arm64.tar.gz/download
Step 2: Unzip the downloaded file
Unzip the downloaded tar.gz file:
Bash:
tar -xzvf x13as-mac-arm64.tar.gz
Step 3: Move the executable to the system path
Move the unzipped x13as executable to the /usr/local/bin directory and make sure it has execute permissions:
Bash:
sudo mv x13as /usr/local/bin/x13as
sudo chmod +x /usr/local/bin/x13as
Step 4: Test the installation
Test that x13as was installed successfully:
Bash:
x13as
If everything went well, we should see the following error message, indicating that the binary worked properly:
information on how to run X-13ARIMA-SEATS.
X-13ARIMA-SEATS Seasonal Adjustment Program
Version Number 1.1 Build 61
Execution began Jul 14, 2022 19.38.34
ERROR: Must specify either an input specification file name
(-i infile or infile) or an input metafile name (-m metafile).
See Section 2 of the X-13ARIMA-SEATS Reference Manual for more
information on how to run X-13ARIMA-SEATS.
Step 5: Use X-13ARIMA-SEATS in Python
We can use the statsmodels library to call X-13ARIMA-SEATS in Python:
import statsmodels.api as sm
import pandas as pd
import numpy as np
# Generate test data
np.random.seed(0)
n_periods = 120 # 10 years, 12 months per year
date_range = pd.date_range(start='2010-01-01', periods=n_periods, freq='M')
seasonal_pattern = np.sin(2 * np.pi * date_range.month / 12)
trend = np.linspace(0, 10, n_periods)
noise = np.random.normal(0, 0.5, n_periods)
data = 10 + trend + seasonal_pattern + noise
# Create time series data
s = pd.Series(data, index=date_range)
# Seasonal Adjustment using X-13ARIMA-SEATS
res = sm.tsa.x13_arima_analysis(s, x12path='/usr/local/bin/x13as')
# Print Results
print(res.summary())
# Plot the Original Data and Seasonally Adjusted Data
import matplotlib.pyplot as plt
fig, ax = plt.subplots(figsize=(12, 8))
s.plot(ax=ax, label='Original Data')
res.seasadj.plot(ax=ax, label='Seasonally Adjusted Data')
plt.legend()
plt.title('Seasonal Adjustment using X-13ARIMA-SEATS')
plt.show()
Out:
With these steps, we should be able to successfully install and use X-13ARIMA-SEATS on macOS.
References: