Why the program does not recognize my variables in egen syntax?

[Phuc Nguyen]

Dear all experts,

I try to calculate the average of some variables for each observation.
My code is:

egen laporta_ave = mean(con_of_corr rul_of_law reg_qua gov_eff poli_stab voice_acc), by(obsnum) /*The "by(obsnum)" option tells Stata to calculate the average for each observation.*/

I am sure the variables in the mean() exist in my daatset. I am wondering why they turn out :

.                  egen laporta_ave = mean(con_of_corr rul_of_law reg_qua gov_eff poli_stab voice_acc), by(obsnum) /*The "by(obsnum)" option tells Stata to calculate the average for each observation.*/
con_of_corrrul_of_lawreg_quagov_effpoli_stabvoice_acc invalid name
r(198);

Thank you so much and best regards

[Joro Kolev]

It is not clear what you want to do here. -egen, mean()- expects a single expression as an argument, you are supplying 3 variables. You might be thinking of -egen, rowmean()-.

[Carlo Lazzaro]

Phuc:
if you have a cross-sectional dataset, the mean of a given variable for a given observation is the value of that variable (if observed).
Are you dealing with a cross-sectional or a panel dataset?
Do you need the mean for different variables for the same observation?
Do you need the mean of a given variables for a given panel?
Please provide interested listers with more details. Thanks.

[Phuc Nguyen]

It is not clear what you want to do here. -egen, mean()- expects a single expression as an argument, you are supplying 3 variables. You might be thinking of -egen, rowmean()-.

Thank you Joro Kolev , I have tried many ways around but not yet fixed it, so I used the conventional approach that I divide sum of 6 variable by 6. Quite less aesthetic but direct, I deem.

Thank you Carlo Lazzaro, regarding my data, it is a panel dataset, I need the mean for various variable of each observation. In general, it similar to getting sum of 6 variables and devide by 6. And I need the mean of each observation in the panel

Thank you so much for taking consideration my question.

[Joro Kolev]

You are not providing a data sample using -dataex- and you are not explaining what you want with a reference to this dataset...

But the way how you explain it, it seems to me that you want

egen mymean = rowmean(v1-v6)

and yes, of course you can also do it manually, as long as you deal with less than 30 variables say.

-egen, rowmean()- is powerful because it can take varlists.

[Phuc Nguyen]

You are not providing a data sample using -dataex- and you are not explaining what you want with a reference to this dataset...

But the way how you explain it, it seems to me that you want

egen mymean = rowmean(v1-v6)

and yes, of course you can also do it manually, as long as you deal with less than 30 variables say.

-egen, rowmean()- is powerful because it can take varlists.

Thank you so much, it is exactly what I am looking for