CODEWITHSUNDEEP
pythonscipy
user131983
user131983

Reputation: 3937

TypeError when using Statsmodels to create Time Series

I am struggling to understand how exactly to use ARIMA within statsmodels.

I am trying to fit an ARIMA model to a set of data I have and am using the same idea as in the answer to this question.

But, I don't know what my endog values, which are the explanatory variables, need to be.

My code is as follows and I get the error: 

TypeError: objfunc() takes exactly 2 arguments (20 given)

in the line:

brute(objfunc, grid, args=(opening_price), finish=None)

I am just passing it the 20 data points I have for this time series and am confused what it expects since this is not right.

def objfunc(order, endog):
    fit = ARIMA(endog, order).fit()
    return fit.aic()

from scipy.optimize import brute
grid = (slice(1, 3, 1), slice(1, 3, 1), slice(1, 3, 1))
brute(objfunc, grid, args=(opening_price), finish=None)

Upvotes: 3

Views: 84

Answers (1)

Warren Weckesser
Warren Weckesser

Reputation: 114976

The following might be a solution. You should really include enough code so that it can be copied and run to reproduce the problem.

In the call to brute, change args=(opening_price) to args=(opening_price,). You don't show what opening_price is, but I assume it is a sequence, and when you write args=(opening_price) (which is equivalent to args=opening_price), the elements of opening_price are expanded into separate arguments when passed to objfunc. The correct form, args=(opening_price,), ensures that args is a tuple containing a single element, opening_price.

Upvotes: 2

Related Questions