Extreme Climate indexes

extract_eventInfo(y::AbstractVector{<:Real}, i_beg::Integer, i_end::Integer;
    index=nothing, goal=1, len_min=1, len2peak=0, ignored...)

Extracts information about an event in a signal.

Examples

julia> y = [0, 1, 2, 3, 2, 1, 0];
julia> extract_eventInfo(y, 2, 6)
(i_beg = 2, i_peak = 4, i_end = 5, len = 4, len_left = 2, len_right = 1, peak = 3)

detect_events(lgl::AbstractVector{Bool}; len_min=1, index=nothing, only_max=false, ignored...)
detect_events(y::AbstractVector{<:Real}, lgl::BitVector;
    len_min=1,
    len2peak=0,
    goal=1,
    index=nothing,
    only_max=false,
    ignored...
)

Detects events in a signal based on a logical vector.

Arguments

• y::AbstractVector{<:Real}: The signal to detect events in.
• lgl::BitVector: A logical vector indicating where events occur in the signal.
• len_min=1: (optional) The minimum length of an event.
• len2peak=0: (optional) The minimum length from the peak to the beginning or end of an event.
• goal=1: (optional), -1 or 1. If 1, find the maximum value in the event; If -1, minimum value used.
• index=nothing: (optional) The index of the signal. If provided, the returned indices will be in the index space.
• only_max=false: (optional) If true, only events with the maximum duration will be returned.
• ignored...: (optional) Ignored arguments.

Returns

An array of named tuples, where each tuple represents an event and has the following fields:

• i_beg::Integer: The index of the beginning of the event.
• i_peak::Integer: The index of the peak of the event.
• i_end::Integer: The index of the end of the event.
• len::Integer: The length of the event.
• len_left::Integer: The length from the beginning of the event to the peak.
• len_right::Integer: The length from the peak to the end of the event.
• peak::Real: The value of the peak of the event.

If only_max is true, only the event with the maximum duration will be returned.

Examples

julia> y = [0, 1, 2, 3, 0, 2, 1, 0];
julia> lgl = y .> 0;
julia> detect_events(lgl)
2-element Vector{NamedTuple{(:i_beg, :i_end, :len), Tuple{Int64, Int64, Int64}}}:
 (i_beg = 2, i_end = 4, len = 3)
 (i_beg = 6, i_end = 7, len = 2)
julia> detect_events(y, lgl)
2-element Vector{NamedTuple{(:i_beg, :i_peak, :i_end, :len, :len_left, :len_right, :peak), NTuple{7, Int64}}}:
 (i_beg = 2, i_peak = 4, i_end = 4, len = 3, len_left = 2, len_right = 0, peak = 3)
 (i_beg = 6, i_peak = 6, i_end = 7, len = 2, len_left = 0, len_right = 1, peak = 2)

HW_index(anorm::AbstractVector; p_left = 0.99)

Compute the HW index for a given anomaly vector anorm.

Arguments

• anorm::AbstractVector: A vector of anomaly scores.
• p_left::Float64=0.99: The probability of false positives.

Returns

A named tuple with the following fields:

• duration::Int: The duration of the anomaly.
• frequency::Int: The number of anomaly events.
• intensity::Float64: The maximum anomaly score.
• volume::Float64: The sum of anomaly scores.
• PR::Float64: The probability of detection.
• FAR::Float64: The false alarm rate.

Example

julia> HW_index([0.1, 0.2, 0.3, 0.2, 0.1, 0, -0.1, 0.1, 0.2, 0.3])
(duration = 9, frequency = 2, intensity = 0.3, volume = 1.5, PR = 89.99999999999993, FAR = 0.9888888888888889)

julia> HW_index([-1, -1])
(duration = 0, frequency = 0, intensity = NaN, volume = NaN, PR = NaN, FAR = NaN)

References

1. Kong, D., Gu, X., Li, J., Ren, G., & Liu, J. (2020). Contributions of Global Warming and Urbanization to the Intensification of Human‐Perceived Heatwaves Over China. Journal of Geophysical Research: Atmospheres, 125(18). https://doi.org/10.1029/2019JD032175