Index

MSModel(ds; k_switching=2, intercept="switching")

Compute the autocorrelation of a dataset ds along the time dimension.

Arguments

• ds: The input dataset.
• lags: The number of lags to compute autocorrelation for. Default is 30.

Returns

The autocorrelation of the input dataset ds along the time dimension.

autocorrelation(ds; lags=30)

Compute the autocorrelation of a dataset ds along the time dimension.

Arguments

• ds: The input dataset.
• lags: The number of lags to compute autocorrelation for. Default is 30.

Returns

The autocorrelation of the input dataset ds along the time dimension.

climato_tp(ds; fct::Function=sum, iduree=3, lead=0)

Calcul de la somme des précipitations pour tous les mois pour la durée spécifié par iduree. En ce moment, l'overlap des années n'est pas pris en charge: si iduree=3, on aura un Cube qui contient les accumulations sur 3 mois pour toutes les séquences de 3 mois, allant de janvier (janvier-février-mars) à octobre (octobre-novembre-décembre).

cumsum(xout, xin)

Compute the cumulative sum of the elements in xin and store the result in xout.

Arguments

• xout: An array to store the cumulative sum.
• xin: An array containing the elements to be summed.

cumsum(cube::YAXArray, kwargs...)

Compute the cumulative sum along the "time" dimension of the input cube.

Arguments

• cube: The input YAXArray cube.
• kwargs: Additional keyword arguments.

Returns

A new YAXArray cube with the cumulative sum computed along the "time" dimension.

daily_fct(xout, xin; fct::Function, index_list = time_to_index)

Apply a function fct to aggregate data from xin into xout on a daily basis used internally

Arguments

• xout: Output array to store the aggregated data.
• xin: Input array containing the data to be aggregated.
• fct: Function to be applied for aggregation.
• index_list: (optional) Function to convert time to index.

daily_fct(cube::YAXArray; fct::Function=mean, shifthour=0, kwargs...)

Apply a function fct to process the data in cube on a daily basis. Assumes that the Cube have sub-daily values.

Arguments

• cube::YAXArray: The input data cube.
• fct::Function: The function to apply for processus sub-daily values. Default is mean.
• shifthour::Int: The number of hours to shift the time values. Default is 0.
• kwargs...: Additional keyword arguments to be passed to the aggregation function.

Returns

A new YAXArray with aggregated data on a daily basis.

diff(xout, xin)

Compute the difference between consecutive elements of xin and store the result in xout. A 0.0 is added at the beginning of the output array.

Arguments

• xout: An array to store the difference values.
• xin: An array of values.

diff(cube::YAXArray, kwargs...)

Compute the difference between consecutive time steps in the given cube.

Arguments

• cube::YAXArray: The input data cube.
• kwargs...: Additional keyword arguments.

Returns

• A new YAXArray object containing the difference between consecutive time steps.

qqmap(dataout, obsvec, refvec, futvec, days, obs_jul, ref_jul, fut_jul; method::String="Additive", detrend::Bool=true, order::Int=4, window::Int64=15, rankn::Int64=50, qmin::Real=0.01, qmax::Real=0.99, thresnan::Float64=0.1, keep_original::Bool=false, interp=Linear(), extrap=Interpolations.Flat())

The qqmap2 function performs quantile mapping bias correction on the futvec data using the obsvec and refvec data as references. It corrects the values in futvec for each day specified in days based on the quantiles estimated from the corresponding days in obsvec and refvec.

Arguments

• dataout: Output array where the corrected values will be stored.
• obsvec: Array of observed values.
• refvec: Array of reference values.
• futvec: Array of future values to be corrected.
• days: Array of days of the year for which correction will be performed.
• obs_jul: Array of Julian day values corresponding to obsvec.
• ref_jul: Array of Julian day values corresponding to refvec.
• fut_jul: Array of Julian day values corresponding to futvec.

Optional Arguments

• method::String: Method used for correction. Default is "Additive".
• detrend::Bool: Whether to detrend the data before correction. Default is true.
• order::Int: Order of the polynomial used for detrending. Default is 4.
• window::Int64: Size of the moving window for selecting reference values. Default is 15.
• rankn::Int64: Number of quantiles to estimate. Default is 50.
• qmin::Real: Minimum quantile value. Default is 0.01.
• qmax::Real: Maximum quantile value. Default is 0.99.
• thresnan::Float64: Threshold for the percentage of NaN values allowed in the data. Default is 0.1.
• keep_original::Bool: Whether to keep the original values in case of too many NaN values. Default is false.
• interp: Interpolation method used for building the correction function. Default is Linear().
• extrap: Extrapolation method used for building the correction function. Default is Interpolations.Flat().

The function modifies dataout in-place and returns nothing.

qqmap(obs::YAXArray, ref::YAXArray, fut::YAXArray; method::String="Additive", detrend::Bool=true, order::Int=4, window::Int=15, rankn::Int=50, qmin::Real=0.01, qmax::Real=0.99, thresnan::Float64=0.1, keep_original::Bool=false, interp=Linear(), extrap=Interpolations.Flat())

This function performs quantile mapping bias correction on data.

Arguments

• obs::YAXArray: The observed data.
• ref::YAXArray: The reference data.
• fut::YAXArray: The future data.
• method::String: The method to apply on raw data for bias correction. Default is "Additive".
• detrend::Bool: Whether to detrend the data before bias correction. Default is true.
• order::Int: The order of the polynomial used for detrending. Default is 4.
• window::Int: The window size (+/- window relative to a given julian day) for calculating the Julian days. Default is 15.
• rankn::Int: The number of quantiles to use for mapping. Default is 50.
• qmin::Real: The minimum quantile value. Default is 0.01.
• qmax::Real: The maximum quantile value. Default is 0.99.
• thresnan::Float64: The threshold for bias correcting a grid based in presence of NaN values. Default is 0.1 (i.e. if there is more than 10% of NaN values, the grid point is not corrected).
• keep_original::Bool: Whether to keep the original data in the output if there is more than the threshold of NaN values. Default is false.
• interp: The interpolation method to use between quantiles. Default is Linear().
• extrap: The extrapolation method to use over qmin and qmax. Default is Interpolations.Flat().

Returns

• A bias-corrected YAXArray.