Several instances have head_age = 999

[prrathi]

From the data info, this is for head ages that are unknown. If I'm grouping the instances by head age and income, should these instances with head_age = 999 be dropped?

[jdebacker]

@prrathi If you are using the a pandas.groupby, it will group by all age values, including this indicator for a missing value. Whether you should drop these values might depend on what you are trying to do, but for most OG-USA calibrations of age-specific values, I'd drop these. Also, given the few observations above age 80, it maybe best to compute age-specific values for ages 20-80 and then we can interpolate values for ages above 80.

[prrathi]

Should the observations above age 80 be combined with age 80, or completely separate? Also same question for the one observation with head age 18 and 36 observations with head age 19

[jdebacker]

@prrathi for consistency at this point, let's not include those over 80 and under 20.