G2

Verified User in Information Technology and Services

Small-Business (50 or fewer emp.)

2/6/2019

"Statsmodels"

4/5

What do you like best about statsmodels?

A comprehensive collection of advanced or special-use statistical and regression functions. Having access to models and statistical tests related to those models in the same package is very handy. Additionally, for regression models such as ARMA/ARIMA/ARIMAX/SARIMAX, the output already contains a lot of relevant information like AIC and BIC score. Review collected by and hosted on G2.com.

What do you dislike about statsmodels?

Output type is inconsistent and can be a hurdle at times. For example some methods will output a list, others a dictionary, etc, which can make using a statsmodels output as an input somewhere else a little tricky. Documentation is also inconsistent. Review collected by and hosted on G2.com.

Recommendations to others considering statsmodels:

Just be aware that outputs from very similar functions aren't necessarily consistent with one another. Some processing is often necessary for use with more standard libraries numpy, pandas, etc. Review collected by and hosted on G2.com.

What problems is statsmodels solving and how is that benefiting you?

I primarily use statsmodels for time series analysis and regression modeling. Having tests such as augmented dickey-fuller and ACF/PACF functions built in are very useful when prepping data for an ARIMA model. Review collected by and hosted on G2.com.