Added diff2 and diff_or_diff2

src/MessyTimeSeries.jl

@@ -20,7 +20,7 @@ module MessyTimeSeries
     # Export methods
     export check_bounds, nan_to_missing!, error_info, verb_message, interpolate_series, forward_backwards_rw_interpolation, centred_moving_average,
             soft_thresholding, solve_discrete_lyapunov, isconverged, trimmed_mean, sum_skipmissing, mean_skipmissing, std_skipmissing, is_vector_in_matrix, 
-            demean, standardise, lag, companion_form, no_combinations, rand_without_replacement, get_bounded_log, get_unbounded_log, get_bounded_logit, get_unbounded_logit;
+            demean, standardise, diff2, diff_or_diff2, lag, companion_form, no_combinations, rand_without_replacement, get_bounded_log, get_unbounded_log, get_bounded_logit, get_unbounded_logit;
 
     # Export functions
     export kfilter!, kfilter_full_sample, kfilter_full_sample!, kforecast, ksmoother, fmin_uc_models, arima, varima, forecast,

src/methods.jl

@@ -353,6 +353,26 @@ standardise(X::FloatMatrix) = (X .- mean(X,dims=2))./std(X,dims=2);
 standardise(X::JVector{Float64}) = (X .- mean_skipmissing(X))./std_skipmissing(X);
 standardise(X::JMatrix{Float64}) = (X .- mean_skipmissing(X))./std_skipmissing(X);
 
+"""
+    diff2(A::AbstractArray, dims::Integer)
+
+Return double-differenced data.
+"""
+diff2(A::AbstractArray; dims::Integer) = diff(diff(A, dims=dims), dims=dims);
+
+"""
+    diff_or_diff2(A::AbstractArray, dims::Integer, use_diff::Bool)
+
+Use either diff or diff2 depending on the value taken by `use_diff`.
+"""
+function diff_or_diff2(A::AbstractArray, dims::Integer, use_diff::Bool)
+    if use_diff
+        return diff(A, dims=dims);
+    else
+        return diff2(A, dims=dims);
+    end
+end
+
 """
     interpolate_series(X::JMatrix{Float64}, n::Int64, T::Int64)