[BUG] The factor varlist feature is not working in some instance of reghdfe

[Lu-Yunzhi]

Bug report

Please complete the following information:

• Stata version: 17.0
• OS: Win 10

Behavior

• Expected behavior: a clear and concise description of what you expected to happen.
  My sample period is from 2008 to 2016. When I set the base group (say, year = 2008) to the year variable, I expect that the group 2008 is treated as baseline and dropped from the possible collinearity problem (since the iteracted variable mX is time-invariant).

• Actual behavior: what was the error. If possible, copy-paste output. For instance:

. reghdfe y ib2008.year##c.mX, absorb(cityid year)

the sample of year = 2016 is dropped

Steps to reproduce the problem

set obs 10
gen id = _n
expand 2016 - 2008 + 1
bys id: gen year = 2008 + _n - 1
gen X = rnormal()
gen y = X*0.5 + rnormal()
bys id (year): egen mX = mean(X)
reghdfe y ib2008.year#c.mX, absorb(id year)