Missing values generated with clock function, in spite of using "double" type

[Smriti Ganapathi]

Dear all,

I have a dataset that looks something like the following, (am trying to calculate duration of the surveys undertaken):

[Clyde Schechter]

Even though you resolved your problem, the etiquette in this community is to leave your post as it was and then add a new post to the thread showing the solution you found. There are a couple of reasons for this.

1. Though it is easy to think of posting here as establishing a dialog with one other person who answers your question, in fact it is a forum with many people silently reading along. Others come periodically and search for a thread on a problem that resembles theirs to find a solution. If we delete problems we have solved, or fail to show the solution, we deprive those members of our community of the opportunity to learn from your solution.

2. If the altruistic motives in #1 do not suffice, it is also possible that if you leave your post up and show your solution, somebody else may come up with a refinement of your solution that will perhaps make it run faster, or would more easily generalize to a different kind of data, or something like that. So you still have an opportunity to learn even if you have solved your own problem.

[Smriti Ganapathi]

Hi Clyde,

You're absolutely right, however I only deleted it because the reason I was facing issues turned out to be quite silly: some of my dates were stored in a different language, so stata's clock function didn't recognise the month and gave me missing values. I wanted to delete the post entirely, but was unable to do so.