Hi everyone,

I am working on the consolidated banking data from the European Central Bank. I set it as panel data (xtset).

I tried to do a regression with more than one independent variable, and received the error r2000 no observations.

I think that is because the data is not tabular.
It consists of rows, each row contains a unique series id, time (quarterly), observation value, item, reference area, etc.
The unique series id is repeated across time.

For example:
ID Country Quarter Item1 Item2 Item3
x_item1 DE Q1 0.5 . .
x_item2 DE Q1 . 1 .
x_item3 DE Q1 . . 5
I think if I combine observations that share an id and a date I will overcome the r2000 error when running the regression.

For example:
Country Quarter Item1 Item2 Item3
DE Q1 0.5 1 5

I tried to reshape the data, but still the problem persists that in every row I have only one observation.

Do you have any suggestions on how to proceed?

I am so grateful for your assistance.

Thank you,
Lydia Gad