I 'm trying to do a data cleaning to a large dataset (3200 variables with 108000 subjects). There are cells with values like " <190" or " >=85" or a text " overweight" at variables with anthropometric measurements. Could you advice me which is the best way to treat these cases in a descriptive analysis. For example, if the variable of interest measures the height and there are 100 cells with values " <150cm" what is better to do? How can i calculate the mean?