In [1]: import numpy as np
In [2]: import statsmodels.api as sm
Generate some data from an ARMA process
In [3]: from statsmodels.tsa.arima_process import arma_generate_sample
In [4]: np.random.seed(12345)
In [5]: arparams = np.array([.75, -.25])
In [6]: maparams = np.array([.65, .35])
The conventions of the arma_generate function require that we specify a 1 for the zero-lag of the AR and MA parameters and that the AR parameters be negated.
In [7]: arparams = np.r_[1, -arparams]
In [8]: maparam = np.r_[1, maparams]
In [9]: nobs = 250
In [10]: y = arma_generate_sample(arparams, maparams, nobs)
Now, optionally, we can add some dates information. For this example, we’ll use a pandas time series.
In [11]: import pandas
In [12]: dates = sm.tsa.datetools.dates_from_range('1980m1', length=nobs)
In [13]: y = pandas.TimeSeries(y, index=dates)
In [14]: arma_mod = sm.tsa.ARMA(y, freq='M')
In [15]: arma_res = arma_mod.fit(order=(2,2), trend='nc', disp=-1)