A Problem of Combining Two Columns in Panel Data

[Minch Park]

Consider the following example describing a panel data set:

clear all
version 18
mata:
N = 100
T = 3
pid  = (1::N) # J(T, 1, 1)
tid  = J(N, 1, 1::T)
V = rnormal(N * T, 1, 0, 1)

nvs  = st_addvar("double", ("pid", "tid", "V"))
st_addobs(length(V) - st_nobs())
st_store(., nvs, (pid, tid, V))
end
xtset pid tid

Using this data, I have generated a matrix containing the first-and second-difference of V.

mata: st_view(dV = ., ., "S" :+ strofreal((1, 2)) :+ "." :+ "V")
mata: dV[1::9, ]
                  1              2
    +-------------------------------+
  1 |             .              .  |
  2 |   1.743666979              .  |
  3 |   1.124789274    2.868456253  |
  4 |             .              .  |
  5 |   -1.05772296              .  |
  6 |   .4311303958   -.6265925644  |
  7 |             .              .  |
  8 |  -.2559339912              .  |
  9 |   1.213522593    .9575886021  |
    +-------------------------------+

Here, I want to combine the two columns without missing values like below:

                  1
    +----------------+
  1 |   1.743666979  |
  2 |   1.124789274  |
  3 |   2.868456253  |
  4 |   -1.05772296  |
  5 |   .4311303958  |
  6 |  -.6265925644  |
  7 |  -.2559339912  |
  8 |   1.213522593  |
  9 |   .9575886021  |
    +----------------+

A solution I found is

mata: Vt = colshape(dV[, ], cols(dV) * 3)
mata: colshape((Vt[, 3], Vt[, 5], Vt[, 6]), 1)

However, I think, there could be a more tidy way to combine the column vectors.
Do you have some good ideas?