This package provides convenient functions to lead&lag vectors with respect to a time vector. The time vector needs to be strictly increasing, but gaps are allowed. This is a common operation when dealing with panel data, where entities may have different missing periods.
The key function in this package is tlag (tlead):
julia> t, v = [1;2;4], [1;2;3];
julia> tlag(t, v) # the default lag period is the unitary difference in t, here 1
3-element Vector{Union{Missing, Int64}}:
missing
1
missing
julia> tlag(t, v, 2) # we can also specify lags using the third argument
3-element Vector{Union{Missing, Int64}}:
missing
missing
2
julia> using Dates;
julia> t = [Date(2020,1,1); Date(2020,1,2); Date(2020,1,4)];
julia> tlag(t, [1, 2, 3]) # customized types of the time vector are also supported
3-element Vector{Union{Missing, Int64}}:
missing
1
missing
julia> tlag(t, [1, 2, 3], Day(2)) # specify two-day lags
3-element Vector{Union{Missing, Int64}}:
missing
missing
2Function tlead shifts the array in the opposite direction, and function tshift calls tlag when the period n is positive and vice versa.
For convenience (and to honor the name of the package), I also define functions panellag, panellead and panelshift to shift vectors in panel data. These functions are wrappers of groupby, transform! and tshift, e.g.,
function panellag!(df, id, t, x, newx, n=oneunit(df[1, t] - df[1, t]); checksorted=true)
return transform!(groupby(df, id), [t, x] => ((t, x) -> tlag(t, x, n; checksorted=checksorted)) => newx)
endIt groups df by id, applies tlag to x with respect to t, and stores the lagged column in df under the name newx.
As an example:
julia> using DataFrames;
julia> df = DataFrame(
t = [1;2;3;4; 1;3;4; 1;4; 1],
id = [1;1;1;1; 2;2;2; 3;3; 4],
x = [1;2;3;4; 5;6;7; 8;9; 10]
);
julia> panellead!(df, :id, :t, :x, :Fx)
10×4 DataFrame
Row │ t id x Fx
│ Int64 Int64 Int64 Int64?
─────┼──────────────────────────────
1 │ 1 1 1 missing
2 │ 2 1 2 1
3 │ 3 1 3 2
⋮ │ ⋮ ⋮ ⋮ ⋮
8 │ 1 3 8 missing
9 │ 4 3 9 missing
10 │ 1 4 10 missing
4 rows omitted