https://github.com/nixtla logo
#neural-forecast
Title
# neural-forecast
d

DAVID FERNANDO PARADA BRIJALBA

08/02/2023, 3:57 PM
Hi everyone, I'm currently working on a macroeconomic time series forecasting project at the Central Bank of Colombia. Our goal is to forecast the IPC (Consumer Price Index) based on the GDP (Gross Domestic Product) data. However, I am facing a challenge in introducing the GDP variable as an exogenous variable since it is observed on a quarterly basis. My question is, how can I handle this issue? thanks 🙂
c

Cristian (Nixtla)

08/02/2023, 5:56 PM
Hi @DAVID FERNANDO PARADA BRIJALBA! The problem is that your target variable has a different sampling frequency, like monthly?
d

DAVID FERNANDO PARADA BRIJALBA

08/02/2023, 6:35 PM
yes, that's right
c

Cristian (Nixtla)

08/02/2023, 6:40 PM
I would recommend to do something like
ffill
pandas function, where you assign the same GDP value to all months of the quarter, and complement this with other calendar exogenous like dummy of month, or month of the quarter (eg. 1, 2, 3).
The input dataframe needs one observation of the GDP for each row/timestamp of your data
d

DAVID FERNANDO PARADA BRIJALBA

08/02/2023, 8:45 PM
Oh, thanks a lot! @Cristian (Nixtla) I'm gonna give it a try 👍🏻.