I would like to know how I can get the probabilities rows (forecast points Lo 80, Hi 80 etc.) for this model:
!pip install pmdarima
import pmdarima as pm
import requests
import numpy as np
import pandas as pd
dataset = pd.DataFrame(
{
"demand": np.array(list(range(1827))),
"kolumn1": np.array(list(range(1828, 3655))),
"kolumn3": np.array(list(range(1828, 3655)))
},
index = pd.date_range("20130101", periods=1827))
lol = dataset.resample('M').mean()
model = pm.auto_arima(lol['demand'],
m=12, seasonal=True,
start_p=0, start_q=0, max_order=4, test='adf',error_action='ignore',
suppress_warnings=True,
stepwise=True, trace=True)
train=lol[(lol.index.get_level_values(0) >= '2018-01-31') & (lol.index.get_level_values(0) <= '2018-06-30')]
test=lol[(lol.index.get_level_values(0) > '2018-06-30')]
test
model.fit(train['demand'])
forecast=model.predict(n_periods=4, return_conf_int=True)
print(summary(forecast))
Unfortunately, after calling the 'summary' function I get this error:
NameError Traceback (most recent call last) in <cell line: 30>() 28 forecastt=model.predict(n_periods=4, return_conf_int=True) 29 ---> 30 print(summary(forecastt))
NameError: name 'summary' is not defined