Mario Forni, Alessandro Giovannelli, Marco Lippi, and Stefano Soccorsi, 
"Dynamic Factor Model with Infinite-Dimensional Factor Space: 
Forecasting", Journal of Applied Econometrics, Vol. 33, No. 5, 2018, pp. 
625-642.

The file fgls-data.zip contains two .xlsx files.

The file fgls-files.zip contains two .csv file, which contain the same
data, as well as two Matlab script files.

#####

DATASET FOR THE US

The Excel sheet DataFGLS.xlsx includes the original data used in the
paper: see Appendix F.

In the first and second rows, the short and the long description of
each series are reported. 

In the third row, we report the code, denoted by Tcode in Appendix F,
indicating the transformation undergone by each series. 

The matlab script LoadinData.m loads the data. The Matlab version used
is R2016b, but it has been also tested on matlab R2014b, and it works
fine.

The series used in the forecasting exercise are 

INDUSTRIAL PRODUCTION,  number 1, referred to as IP in the paper, 

CONSUMER PRICE INDEX, CPIAUCL,  number 77, reported to as CPI in the paper.

#####

DATASET FOR THE EURO AREA

The Excel sheet DataEUFGLS.xlsx includes the original data used in the
paper: see Appendix G.

In the first and second rows, the short and the long description of
each series are reported. 

In the third row, we report the code, denoted by Tcode in Appendix G,
indicating the transformation undergone by each series. 

The matlab script LoadinDataEU.m loads the data. The Matlab version
used is R2016b, but it has been also tested on matlab R2014b ,and it
works fine.

The series used in the forecasting exercise are 

INDUSTRIAL PRODUCTION, EK PRODUCTION - TOTAL INDUSTRY EXCL.
CONSTRUCTION, number 93, referred to as IP in the paper, 

CONSUMER PRICE INDEX, EM CPI,  number 69, reported to as CPI in the paper.