dealing with no never treated units in data

[marcelortizv]

This PR supersedes PR#220 and implements two minor improvements in did:

1. Handling Missing Never-Treated Groups:
   When the data contain no never-treated units but the user specifies "nevertreated" as the control group, did now issues a warning instead of throwing an error. In this situation, the last treated cohort (as defined by gname) is used as a “synthetic” never-treated group. We then drop all observations from periods ≥ latest_g. This behavior applies in both the regular and fast estimation modes. Unit tests verify that a warning is raised and that both modes produce identical treatment‐effect estimates.

Unit test in test-att_gt.R, Lines 215-226

# try to use never treated group as comparison group, should warn
expect_warning(nonev_orig <- att_gt(yname="Y", xformla=~X, data=data, tname="period",
                    control_group="nevertreated",
                    gname="G", est_method="dr", panel=FALSE, faster_mode = FALSE))

# try to use never treated group as comparison group with faster mode, should warn
expect_warning(nonev_faster <- att_gt(yname="Y", xformla=~X, data=data, tname="period",
                      control_group="nevertreated",
                      gname="G", est_method="dr", panel=FALSE, faster_mode = TRUE))

# make sure both methods give same ATT(g,t) with no never treated group
expect_equal(nonev_orig$att, nonev_faster$att)

2. Suppressing Warnings for Unbalanced Panels/RCS:
   To prevent spurious warnings in unbalanced panels or repeated cross‐section settings, I’ve added a fill = TRUE argument when constructing the influence‐function matrix. The warning previously read:

Column 3 ['value'] of item 2 is missing in item 1. Use fill=TRUE to fill with NA (NULL for list columns), or use.names=FALSE to ignore column names. use.names='check' (default from v1.12.2) emits this message and proceeds as if use.names=FALSE for backwards compatibility. See news item 5 in v1.12.2 for options to control this message.

All unit tests are passing from my end