Python OLS model: __init__() missing 1 required positional argument: 'endog'

Question

import statsmodels.api as sm
xdat = rets['EUROSTOXX']
xdat = sm.add_constant(xdat)
ydat = rets['VSTOXX']
model = sm.OLS(y=ydat,x=xdat).fit()

I don't understand why there is an error popping up as stated on the topic. The following is the tail of the Dataframe of rets

Out[105]:
            EUROSTOXX   VSTOXX
2014-12-23  0.011835    -0.039307
2014-12-24  -0.002449   0.000000
2014-12-29  0.000160    0.121598
2014-12-30  -0.015574   0.048998
2014-12-31  0.003336    0.000000

Contredanse

Answer 1

According to the documentation of OLS, the function's parameters aren't called y and x, but endog and exog. You can simply change your function call to:

model = sm.OLS(ydat, xdat).fit()

or:

model = sm.OLS(endog=ydat, exog=xdat).fit()

Tiphanie