Creating new variables using the subset of an existing variable

[Volkan Arslan]

Hello,

I am pretty new user of Stata. I would like to create new variables by using the subset of an existing variable. I have the values of the first 5 columns of the table stated below. I want to create the last three columns.
Analyst A released her forecasts for EPS1, EPS2, and EPS3 on March 2014. We can see these values at the fifth column. I would like to create the last three new columns by matching the name of analyst from the first column and also check Date of forecasting and forecast date. It seems a bit confusing for me.

Thanks in advance for your help.

Volkan

Analyst	FPI	Forecast Date	Date of forecasting	Forecasted Value (EPS)	EPS1 (forecast for year 1 end)	EPS2 (forecast for year 2 end)	EPS3 (forecast for year 3 end)
A	1	2014 December	March 2014	1.73 (it is EPS1 since FPI=1)	1.73	1.89	2.25
A	2	2015 December	March 2014	1.89 (EPS2 it is EPS2 since FPI=2)
A	3	2016 December	March 2014	2.25 (EPS3)
B	1	2014 December	April 2014	1.85	1.85

[Nick Cox]

sort Analyst FPI 
forval j = 1/3 { 
      by Analyst : gen EPS`j' = EPS[`j'] 
}

[Volkan Arslan]

Thanks Nick. It works. Can I add more criteria into 'sort' function? For example if analyst has also forecasts for different companies

Analyst	FPI	Company	Forecast Date	Date of forecasting	Forecasted Value (EPS)	EPS1 (forecast for year 1 end)	EPS2 (forecast for year 2 end)	EPS3 (forecast for year 3 end)
A	1	Apple	2014 December	March 2014	1.73 (it is EPS1 since FPI=1)	1.73	1.89	2.25
A	2	Apple	2015 December	March 2014	1.89 (EPS2 it is EPS2 since FPI=2)	1.73	1.89	2.25
A	3	Apple	2016 December	March 2014	2.25 (EPS3)	1.73	1.89	2.25
A	1	Orange	2014 December	April 2014	1,46	1.46	1.76
A	2	Orange	2015 December	April 2014	1.76	1.46	1.76
B	1	Peach	2014 December	April 2014	1.85	1.85

[Volkan Arslan]

Is there any practical way of reorganizing data? I have values in one column. I just wanted to create additional columns based on this column.