# Thread: Dealing with censored dates as regressors

1. ## Dealing with censored dates as regressors

I'm estimating a linear model, with one of the independent variables being a date of appearance. I consider the date to be a minimum date of appearance, not an obeserved one. Thus it is censored. What would be the proper estimation to approach this?

2. ## Re: Dealing with censored dates as regressors

I do not follow what you are exactly doing. Can you add some more information. So using linear regression and what exactly is this term you are referencing and how are you changing it?

3. ## Re: Dealing with censored dates as regressors

Originally Posted by hlsmith
I do not follow what you are exactly doing. Can you add some more information. So using linear regression and what exactly is this term you are referencing and how are you changing it?
I'm using a linear model in which the dependent variable is rental price for apartment i in month t, and one of the regressors is a counting variable of some sort, for apartment i in time t. The problem is that for me, the counting in time t functions as a minimum value, and not the exact count at that time. I'm looking for an estimation to treat a variable as a minimum value for each observation.

 Tweet