Summarizing Float64 values after grouping by different time scales from DateTime object

General Usage
, ,
2

I am assuming by looking at the R code that you want to compute summary of values every hour? It is a timeseries problem. You can use TSFrames.jl:

julia> using TSFrames

julia> time = [
    DateTime(2017, 01, 01, 0, 0),
    DateTime(2017, 01, 01, 0, 20),
    DateTime(2017, 01, 01, 0, 40),
    DateTime(2017, 01, 01, 01, 00),
    DateTime(2017, 01, 01, 01, 20),
    DateTime(2017, 01, 01, 01, 40)
];
julia> values = [100, 200, 300, 500, 600, 700];

julia> ts = TSFrame(values, time)
TSFrame(values, time)
6×1 TSFrame with DateTime Index
 Index                x1
 DateTime             Int64
────────────────────────────
 2017-01-01T00:00:00    100
 2017-01-01T00:20:00    200
 2017-01-01T00:40:00    300
 2017-01-01T01:00:00    500
 2017-01-01T01:20:00    600
 2017-01-01T01:40:00    700

## compute hourly mean
julia> apply(ts, Hour(1), mean)
2×1 TSFrame with DateTime Index
 Index                x1_mean
 DateTime             Float64
──────────────────────────────
 2017-01-01T00:00:00    200.0
 2017-01-01T01:00:00    600.0

Here, Hour(1) specifies the period to summarise on and mean is the summary function.

REF: TSFrames API docs