Python Forum

Hello,

I am trying to get the residuals from the AutoReg function from StatsModels but it outputs NaNs:

from statsmodels.tsa.ar_model import AutoReg

df = stock['Close'] / stock['Close'].shift(1) - 1
df = df.dropna()

model = AutoReg(df, lags=1, old_names=False).fit()
res_values = model.resid

res_values

Output:
Out[12]: 
Date
2010-03-05   NaN
2010-03-09   NaN
2010-03-10   NaN
2010-03-11   NaN
2010-03-12   NaN
              ..
2021-04-02   NaN
2021-04-05   NaN
2021-04-06   NaN
2021-04-07   NaN
2021-04-08   NaN
Length: 2775, dtype: float64

I'm not proficient in StatsModels or statistics so I'm sure my mistake is novice but I can't seem to find an answer why.

Matt