evalhyd::evalp

template<class XD2, class XD4, class XB4 = xt::xtensor<bool, 4>, class XS2 = xt::xtensor<std::array<char, 32>, 2>>
std::vector<xt::xarray<double>> evalhyd::evalp(const xt::xexpression<XD2> &q_obs, const xt::xexpression<XD4> &q_prd, const std::vector<std::string> &metrics, const xt::xexpression<XD2> &q_thr = XD2({}), xtl::xoptional<const std::string, bool> events = xtl::missing<const std::string>(), const std::vector<double> &c_lvl = {}, const xt::xexpression<XB4> &t_msk = XB4({}), const xt::xexpression<XS2> &m_cdt = XS2({}), xtl::xoptional<const std::unordered_map<std::string, int>, bool> bootstrap = xtl::missing<const std::unordered_map<std::string, int>>(), const std::vector<std::string> &dts = {}, xtl::xoptional<const int, bool> seed = xtl::missing<const int>(), xtl::xoptional<const std::vector<std::string>, bool> diagnostics = xtl::missing<const std::vector<std::string>>())

Function to evaluate probabilistic streamflow predictions.

Template Parameters

XD2: Any 2-dimensional container class storing numeric elements

(e.g. xt::xtensor<double, 2>, xt::pytensor<double, 2>, xt::rtensor<double, 2>, etc.).

XD4: Any 4-dimensional container class storing numeric elements

(e.g. xt::xtensor<double, 4>, xt::pytensor<double, 4>, xt::rtensor<double, 4>, etc.).

XB4: Any 4-dimensional container class storing boolean elements

(e.g. xt::xtensor<bool, 4>, xt::pytensor<bool, 4>, xt::rtensor<bool, 4>, etc.).

XS2: Any 2-dimensional container class storing string elements

(e.g. xt::xtensor<std::array<char, 32>, 2>, xt::pytensor<std::array<char, 32>, 2>, xt::rtensor<std::array<char, 32>, 2>, etc.).

Parameters

q_obs: XD2

Streamflow observations. Time steps with missing observations must be assigned NAN values. Those time steps will be ignored both in the observations and the predictions before the metrics are computed. shape: (sites, time)

q_prd: XD4

Streamflow predictions. Time steps with missing predictions must be assigned NAN values. Those time steps will be ignored both in the observations and the predictions before the metrics are computed. shape: (sites, lead times, members, time)

metrics: std::vector<std::string>

The sequence of evaluation metrics to be computed.

See also

Probabilistic metrics

q_thr: XD2, optional

Streamflow exceedance threshold(s). If provided, events must also be provided. shape: (sites, thresholds)

events: std::string, optional

The type of streamflow events to consider for threshold exceedance-based metrics. It can either be set as “high” when flooding conditions/high flow events are evaluated (i.e. event occurring when streamflow goes above threshold) or as “low” when drought conditions/low flow events are evaluated (i.e. event occurring when streamflow goes below threshold). It must be provided if q_thr is provided.

c_lvl: std::vector<double>, optional

Confidence interval(s).

t_msk: XB4, optional

Mask(s) used to generate temporal subsets of the whole streamflow time series (where True/False is used for the time steps to include/discard in a given subset). If not provided and neither is m_cdt, no subset is performed and only one set of metrics is returned corresponding to the whole time series. If provided, as many sets of metrics are returned as they are masks provided. shape: (sites, lead times, subsets, time)

See also

Temporal masking

m_cdt: XS2, optional

Masking conditions to use to generate temporal subsets. Each condition consists in a string and can be specified on observed/predicted streamflow values/statistics (mean, median, quantile), or on time indices. If provided in combination with t_msk, the latter takes precedence. If not provided and neither is t_msk, no subset is performed and only one set of metrics is returned corresponding to the whole time series. shape: (sites, subsets)

See also

Conditional masking

bootstrap: std::unordered_map<std::string, int>, optional

Parameters for the bootstrapping method used to estimate the sampling uncertainty in the evaluation of the predictions. The parameters are: ‘n_samples’ the number of random samples, ‘len_sample’ the length of one sample in number of years, and ‘summary’ the statistics to return to characterise the sampling distribution). If not provided, no bootstrapping is performed. If provided, dts must also be provided.

See also

Bootstrapping

dts: std::vector<std::string>, optional

Datetimes. The corresponding date and time for the temporal dimension of the streamflow observations and predictions. The date and time must be specified in a string following the ISO 8601-1:2019 standard, i.e. “YYYY-MM-DD hh:mm:ss” (e.g. the 21st of May 2007 at 4 in the afternoon is “2007-05-21 16:00:00”). The time series must feature complete years. Only minute, hourly, and daily time steps are supported. If provided, it is only used if bootstrap is also provided.

seed: int, optional

A value for the seed used by random generators. This parameter guarantees the reproducibility of the metric values between calls.

diagnostics: std::vector<std::string>, optional

The sequence of evaluation diagnostics to be computed.

See also

Diagnostics

Returns

std::vector<xt::xarray<double>>

The sequence of evaluation metrics computed in the same order as given in metrics, followed by the sequence of evaluation diagnostics computed in the same order as given in diagnostics. shape: (metrics+diagnostics,)<(sites, lead times, subsets, samples, {quantiles,} {thresholds,} {components,} {ranks,} {intervals})>

Examples

#include <xtensor/xtensor.hpp>
#include <evalhyd/evalp.hpp>

xt::xtensor<double, 2> obs = {{ 4.7, 4.3, 5.5, 2.7, 4.1 }};
xt::xtensor<double, 4> prd = {{{{ 5.3, 4.2, 5.7, 2.3, 3.1 },
                                { 4.3, 4.2, 4.7, 4.3, 3.3 },
                                { 5.3, 5.2, 5.7, 2.3, 3.9 }}}};
xt::xtensor<double, 2> thr = {{ 4.7, 4.3, 5.5, 2.7, 4.1 }};

evalhyd::evalp(obs, prd, {"BS"}, thr);

xt::xtensor<bool, 3> msk = {{{ false, true, true, false, true }}};

evalhyd::evalp(obs, prd, {"BS"}, thr, msk);

evalhyd::evalp(obs, prd, {"CRPS_FROM_QS"});