Data Values Showing Not Equal When Actually Equal

[Zachariah Rutledge]

I am working on a wage analysis where I created a duplicate variable (called wage2) that is equal to a variable called wage. I saved the wage2 variable as it's own data set along with the individual ID variable and merged it into a different data set that contains the variable wage based on individual ID. I saved wage2 as a quality control check to check it against the wage variable. However, when I ask Stata to count the observations where "wage" is not equal to "wage2", it tells me there are 18 observations that don't match (out of a total of 133). However, when I visually inspect the data, the values for the 18 nonmatching entries are identical. Can someone please tell me why Stata is saying that they are not equal when they actually are? The values for the 18 observations are listed below, which I copied from the data browser. Also, the summary statistics for these 18 observations for the wage and wage2 variables are identical.

wage wage2
28.26 28.26
17.41 17.41
31.76 31.76
17.41 17.41
17.41 17.41
28.26 28.26
17.41 17.41
17.41 17.41
17.41 17.41
28.26 28.26
15.97 15.97
14.49 14.49
17.41 17.41
14.49 14.49
17.41 17.41
17.4 17.4
17.41 17.41
28.26 28.26

[Martyn Sherriff]

Have a look at the FAQ at:

https://www.stata.com/support/faqs/d...-point-values/

[Rich Goldstein]

help precision

[Zachariah Rutledge]

Thanks for the help. When I count the observations using the "float" command, the issue goes away.