Logit Error r(2000) - No Observations

[Ellen Tsai]

Hi everyone,

I am new to Stata and have been struggling with running logit. I am using StataSE 13. When I run logit, I keep getting "no observations; r(2000)". Here is basic info about my data set:

Number of observations: 88
Dependent variable: "forsames"
Independent variable: "polaff" (politicalaffiliation)

*Tabulating the dependent variable
tabulate forsames

encode politicalaffiliation, generate (polaff)

tabulate polaff
tabulate polaff, nolabel

*Generating political affiliation dummy variables
generate amindep=0
replace amindep=1 if polaff==1

generate democrat=0
replace democrat=1 if polaff==2

generate green=0
replace green=1 if polaff==3

generate indep=0
replace indep=1 if polaff==4

generate libert=0
replace libert=1 if polaff==5

generate peace=0
replace peace=1 if polaff==6

generate repub=0
replace repub=1 if polaff==7

*Political affiliation as explanatory variable, democrats as baseline
logit forsames amindep green indep libert peace repub

I have checked the manual and other help threads (e.g. I am using 0,1 variables; I'm not missing any data, etc.), but nothing seems to be working.

I would be incredibly grateful for any help or suggestions!

[Carlo Lazzaro]

Ellen:
welcome to the list.
You may want to post your .dta file so that other listers may want to repeat your analysis.

[Ellen Tsai]

Thank you for the suggestion, Carlo. I have added the .dta file to my original post.

[Carlo Lazzaro]

Ellen:
unfortunately, your attachment was probably lost along the wire.

[Ellen Tsai]

I'm afraid I don't understand what you mean. The .dta file is working for me when I open it.

[Svend Juul]

forsames is a string variable. Look at:

describe forsames

You need a numeric variable:

destring forsames , generate(forsames_n)

[Ellen Tsai]

Oh, I see! Thank you very much for your help. Problem solved

[Carlo Lazzaro]

Ellen:
now that I can open the .dta I second in full Svend's helpful advice.