search

mboeck11 / BGVAR

Toolbox for the estimation of Bayesian Global Vector Autoregressions in R.
27 stars 20 forks source link

Issues starting a model with seemingly valid data. #17

Closed csupremacy closed 1 year ago

csupremacy commented 1 year ago

I have not been able to initiate the bgvar object.

I have Data, weights and exogenous variables dataframes - all read from .xlsx files with readxl. All the entities have two-letter names, white the variables have anywhere between 2 and 8 character names, e.g. : UK.EX, US. NGAS_Q.

However, initiating a bgvar() object with this data yields

Error in bgvar(Data = dfdata, W = Weights) : 
  Please provide entity names with exactly two characters.

I have tried to create a sample of that data with only a couple entities and variables (just selected only some of the columns and an appropriate subset of the weight matrix and renormalizing the rows to 1). The same error is thrown.

Please find attached the sample of the data.

Am I doing something wrong? Is there an issue of some sort with anything in the model? I do my data wrangling in Python. Could this cause any problems? I don't see any potential solutions in the documentation, not to mention the other publications on this package.

Any hint would be appreciated. trialnovd.xlsx W_novo.xlsx trialexog.xlsx

csupremacy commented 1 year ago

Importing the same data written into a .csv file gives Error in if (any(x == "list(NULL)blabla")) NULL else x : missing value where TRUE/FALSE needed

I don't understand from the source code what this is supposed to mean.