Retime

The retime function allows you to retime, i.e. change the timestamps of a TimeArray, similar to what Matlab's retime does.

using Plots, Dates, TimeSeries
gr()
timestamps = range(DateTime(2020, 1, 1), length = 7*24, step = Hour(1))
ta = TimeArray(timestamps, cumsum(randn(7*24)), [:a])

168×1 TimeArray{Float64, 1, DateTime, Vector{Float64}} 2020-01-01T00:00:00 to 2020-01-07T23:00:00
┌─────────────────────┬─────────┐
│                     │ a       │
├─────────────────────┼─────────┤
│ 2020-01-01T00:00:00 │ 2.31198 │
│ 2020-01-01T01:00:00 │ 3.87913 │
│ 2020-01-01T02:00:00 │ 3.28199 │
│ 2020-01-01T03:00:00 │  2.2734 │
│ 2020-01-01T04:00:00 │ 2.58031 │
│ 2020-01-01T05:00:00 │ 2.95722 │
│ 2020-01-01T06:00:00 │ 2.87577 │
│ 2020-01-01T07:00:00 │ 4.00162 │
│          ⋮          │    ⋮    │
│ 2020-01-07T17:00:00 │ 7.71878 │
│ 2020-01-07T18:00:00 │ 8.16602 │
│ 2020-01-07T19:00:00 │ 10.0477 │
│ 2020-01-07T20:00:00 │ 9.26863 │
│ 2020-01-07T21:00:00 │ 9.72699 │
│ 2020-01-07T22:00:00 │ 9.93005 │
│ 2020-01-07T23:00:00 │ 11.3293 │
└─────────────────────┴─────────┘
                 153 rows omitted

Using a new time step

retime(ta, Minute(15))

669×1 TimeArray{Float64, 2, DateTime, Matrix{Float64}} 2020-01-01T00:00:00 to 2020-01-07T23:00:00
┌─────────────────────┬─────────┐
│                     │ a       │
├─────────────────────┼─────────┤
│ 2020-01-01T00:00:00 │ 2.31198 │
│ 2020-01-01T00:15:00 │ 2.31198 │
│ 2020-01-01T00:30:00 │ 2.31198 │
│ 2020-01-01T00:45:00 │ 2.31198 │
│ 2020-01-01T01:00:00 │ 3.87913 │
│ 2020-01-01T01:15:00 │ 3.87913 │
│ 2020-01-01T01:30:00 │ 3.87913 │
│ 2020-01-01T01:45:00 │ 3.87913 │
│          ⋮          │    ⋮    │
│ 2020-01-07T21:30:00 │ 9.72699 │
│ 2020-01-07T21:45:00 │ 9.72699 │
│ 2020-01-07T22:00:00 │ 9.93005 │
│ 2020-01-07T22:15:00 │ 9.93005 │
│ 2020-01-07T22:30:00 │ 9.93005 │
│ 2020-01-07T22:45:00 │ 9.93005 │
│ 2020-01-07T23:00:00 │ 11.3293 │
└─────────────────────┴─────────┘
                 654 rows omitted

Using new timestep vector

new_timestamps = range(DateTime(2020, 1, 1), DateTime(2020, 1, 2), step = Minute(15))
retime(ta, new_timestamps)

97×1 TimeArray{Float64, 2, DateTime, Matrix{Float64}} 2020-01-01T00:00:00 to 2020-01-02T00:00:00
┌─────────────────────┬───────────┐
│                     │ a         │
├─────────────────────┼───────────┤
│ 2020-01-01T00:00:00 │   2.31198 │
│ 2020-01-01T00:15:00 │   2.31198 │
│ 2020-01-01T00:30:00 │   2.31198 │
│ 2020-01-01T00:45:00 │   2.31198 │
│ 2020-01-01T01:00:00 │   3.87913 │
│ 2020-01-01T01:15:00 │   3.87913 │
│ 2020-01-01T01:30:00 │   3.87913 │
│ 2020-01-01T01:45:00 │   3.87913 │
│          ⋮          │     ⋮     │
│ 2020-01-01T22:30:00 │   0.84457 │
│ 2020-01-01T22:45:00 │   0.84457 │
│ 2020-01-01T23:00:00 │  -0.39346 │
│ 2020-01-01T23:15:00 │  -0.39346 │
│ 2020-01-01T23:30:00 │  -0.39346 │
│ 2020-01-01T23:45:00 │  -0.39346 │
│ 2020-01-02T00:00:00 │ -0.848762 │
└─────────────────────┴───────────┘
                    82 rows omitted

You can perform retime on irregularly spaced timestamps, both using a TimeArray with irregular timestamps or using a vector of irregular timestamps. Depending on the timestamps upsampling or downsampling is used.

new_timestamps = vcat(
    range(DateTime(2020, 1, 1), DateTime(2020, 1, 2)-Minute(15), step = Minute(15)),
    range(DateTime(2020, 1, 2), DateTime(2020, 1, 3), step = Hour(1)),
)
retime(ta, new_timestamps)

121×1 TimeArray{Float64, 2, DateTime, Matrix{Float64}} 2020-01-01T00:00:00 to 2020-01-03T00:00:00
┌─────────────────────┬─────────┐
│                     │ a       │
├─────────────────────┼─────────┤
│ 2020-01-01T00:00:00 │ 2.31198 │
│ 2020-01-01T00:15:00 │ 2.31198 │
│ 2020-01-01T00:30:00 │ 2.31198 │
│ 2020-01-01T00:45:00 │ 2.31198 │
│ 2020-01-01T01:00:00 │ 3.87913 │
│ 2020-01-01T01:15:00 │ 3.87913 │
│ 2020-01-01T01:30:00 │ 3.87913 │
│ 2020-01-01T01:45:00 │ 3.87913 │
│          ⋮          │    ⋮    │
│ 2020-01-02T18:00:00 │ 1.81283 │
│ 2020-01-02T19:00:00 │ 1.91873 │
│ 2020-01-02T20:00:00 │ 2.60321 │
│ 2020-01-02T21:00:00 │ 3.91545 │
│ 2020-01-02T22:00:00 │ 3.05077 │
│ 2020-01-02T23:00:00 │ 2.98382 │
│ 2020-01-03T00:00:00 │ 2.56973 │
└─────────────────────┴─────────┘
                 106 rows omitted

Upsampling

Interpolation is done using the upsample argument. If no data is directly hit, the specified upsample method is used. Available upsample methods are:

Linear() or :linear
Nearest() or :nearest
Previous() or :previous
Next() or :next

ta_ = retime(ta, Minute(15), upsample=Linear())

669×1 TimeArray{Float64, 2, DateTime, Matrix{Float64}} 2020-01-01T00:00:00 to 2020-01-07T23:00:00
┌─────────────────────┬─────────┐
│                     │ a       │
├─────────────────────┼─────────┤
│ 2020-01-01T00:00:00 │ 2.31198 │
│ 2020-01-01T00:15:00 │ 2.70377 │
│ 2020-01-01T00:30:00 │ 3.09555 │
│ 2020-01-01T00:45:00 │ 3.48734 │
│ 2020-01-01T01:00:00 │ 3.87913 │
│ 2020-01-01T01:15:00 │ 3.72984 │
│ 2020-01-01T01:30:00 │ 3.58056 │
│ 2020-01-01T01:45:00 │ 3.43128 │
│          ⋮          │    ⋮    │
│ 2020-01-07T21:30:00 │ 9.82852 │
│ 2020-01-07T21:45:00 │ 9.87929 │
│ 2020-01-07T22:00:00 │ 9.93005 │
│ 2020-01-07T22:15:00 │ 10.2799 │
│ 2020-01-07T22:30:00 │ 10.6297 │
│ 2020-01-07T22:45:00 │ 10.9795 │
│ 2020-01-07T23:00:00 │ 11.3293 │
└─────────────────────┴─────────┘
                 654 rows omitted

plot(ta)
plot!(ta_)

Downsampling

Downsampling or aggregation is done using the downsample argument. This applies a function to each interval not including the right-edge of the interval. If no data is present in the interval the specified upsample method is used. Available downsample methods are:

Mean() or :mean
Min() or :min
Max() or :max
Count() or :count
Sum() or :sum
Median() or :median
First() or :first
Last() or :last

ta_ = retime(ta, Hour(6), downsample=Mean())

28×1 TimeArray{Float64, 2, DateTime, Matrix{Float64}} 2020-01-01T00:00:00 to 2020-01-07T18:00:00
┌─────────────────────┬───────────┐
│                     │ a         │
├─────────────────────┼───────────┤
│ 2020-01-01T00:00:00 │   2.88067 │
│ 2020-01-01T06:00:00 │   3.80897 │
│ 2020-01-01T12:00:00 │   2.65148 │
│ 2020-01-01T18:00:00 │   1.30297 │
│ 2020-01-02T00:00:00 │  -3.02952 │
│ 2020-01-02T06:00:00 │  -3.68937 │
│ 2020-01-02T12:00:00 │ -0.817812 │
│ 2020-01-02T18:00:00 │   2.71414 │
│          ⋮          │     ⋮     │
│ 2020-01-06T06:00:00 │   8.14749 │
│ 2020-01-06T12:00:00 │   9.15137 │
│ 2020-01-06T18:00:00 │   8.48118 │
│ 2020-01-07T00:00:00 │    7.6436 │
│ 2020-01-07T06:00:00 │   8.07861 │
│ 2020-01-07T12:00:00 │   8.59339 │
│ 2020-01-07T18:00:00 │   9.74477 │
└─────────────────────┴───────────┘
                    13 rows omitted

plot(ta)
plot!(ta_)

Extrapolation

Extrapolation at the beginning and end of the time series is done using the extrapolate argument. Available extrapolate methods are:

FillConstant(value) or :fillconstant
NearestExtrapolate() or :nearest
MissingExtrapolate() or :missing
NaNExtrapolate() or :nan

new_timestamps = range(DateTime(2019, 12, 31), DateTime(2020, 1, 2), step = Minute(15))
ta_ = retime(ta, new_timestamps, extrapolate=MissingExtrapolate())