Skip to content
GitLab
Commit db82a4d1 authored by Thibault Hallouin's avatar Thibault Hallouin
Browse files

use new evalhyd templated functions

parent 5b8c9ca2
main 7-update-xtensor-headers dev v0.1.2.0 v0.1.1.0 v0.1.0.0
1 merge request!1release v0.1.0.0
Hide whitespace changes
Inline Side-by-side
Showing
with 25 additions and 286 deletions
+25 -286
setup.py
+ 1
− 4
View file @ db82a4d1
......@@ -13,10 +13,7 @@ __version__ = '0.0.1'
ext_modules = [
Pybind11Extension(
"evalhyd",
['src/evalhyd-python.cpp',
'deps/evalhyd/src/probabilist/evaluator_brier.cpp',
'deps/evalhyd/src/probabilist/evaluator_elements.cpp',
'deps/evalhyd/src/probabilist/evaluator_quantiles.cpp'],
['src/evalhyd-python.cpp'],
include_dirs=[
numpy.get_include(),
os.path.join(os.getcwd(), 'deps', 'evalhyd', 'deps', 'xtl',
......
src/evalhyd-python.cpp
+ 24
− 282
View file @ db82a4d1
......@@ -6,6 +6,7 @@
#define MACRO_STRINGIFY(x) STRINGIFY(x)
#define FORCE_IMPORT_ARRAY
#include <xtensor/xexpression.hpp>
#include <xtensor/xview.hpp>
#include <xtensor-python/pytensor.hpp>
......@@ -17,22 +18,21 @@ using namespace py::literals;
// reshape 1D tensors to 2D tensors
auto evald_1d(
const xt::xtensor<double, 1>& q_obs,
const xt::xtensor<double, 1>& q_prd,
const xt::pytensor<double, 1>& q_obs,
const xt::pytensor<double, 1>& q_prd,
const std::vector<std::string>& metrics,
const std::string& transform = "none",
const double exponent = 1,
double epsilon = -9,
const xt::xtensor<bool, 2>& t_msk = {},
const xt::xtensor<std::array<char, 32>, 1>& m_cdt = {},
const std::unordered_map<std::string, int>& bootstrap =
{{"n_samples", -9}, {"len_sample", -9}, {"summary", 0}},
const std::vector<std::string>& dts = {}
const std::string& transform,
const double exponent,
double epsilon,
const xt::pytensor<bool, 2>& t_msk,
const xt::pytensor<std::array<char, 32>, 1>& m_cdt,
const std::unordered_map<std::string, int>& bootstrap,
const std::vector<std::string>& dts
)
{
return evalhyd::evald(
xt::view(q_obs, xt::newaxis(), xt::all()),
xt::view(q_prd, xt::newaxis(), xt::all()),
return evalhyd::evald<xt::pytensor<double, 2>, xt::pytensor<bool, 2>>(
xt::pytensor<double, 2>(xt::view(q_obs, xt::newaxis(), xt::all())),
xt::pytensor<double, 2>(xt::view(q_prd, xt::newaxis(), xt::all())),
metrics,
transform,
exponent,
......@@ -49,105 +49,13 @@ PYBIND11_MODULE(evalhyd, m)
{
xt::import_numpy();
m.doc() = R"pbdoc(
Utility for evaluation of streamflow predictions.
)pbdoc";
m.doc() = "Utility for evaluation of streamflow predictions";
// deterministic evaluation
m.def(
"evald", evald_1d,
R"pbdoc(
Function to evaluate deterministic streamflow predictions.
:Parameters:
q_obs: `numpy.ndarray`
1D array of streamflow observations. Time steps with
missing observations must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (time,)
q_prd: `numpy.ndarray`
1D array of streamflow predictions. Time steps with
missing predictions must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (time,)
metrics: `List[str]`
The sequence of evaluation metrics to be computed.
transform: `str`, optional
The transformation to apply to both streamflow observations
and predictions prior to the calculation of the *metrics*.
exponent: `float`, optional
The value of the exponent n to use when the *transform* is
the power function. If not provided (or set to default value
1), the streamflow observations and predictions remain
untransformed.
epsilon: `float`, optional
The value of the small constant ε to add to both the
streamflow observations and predictions prior to the
calculation of the *metrics* when the *transform* is the
reciprocal function, the natural logarithm, or the power
function with a negative exponent (since none are defined
for 0). If not provided (or set to default value -9),
one hundredth of the mean of the streamflow observations
is used as value for epsilon.
t_msk: `numpy.ndarray`, optional
1D array of 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. If
provided, masks must feature the same number of dimensions as
observations and predictions, and it must broadcastable with
both of them.
shape: (subsets, time)
m_cdt: `numpy.ndarray`, optional
1D array of masking condition(s) to use to generate
temporal subsets. Each condition consists in a string and
can be specified on observed 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. If provided, only one condition per time series
of observations can be provided.
shape: (subsets,)
bootstrap: `dict`, optional
Parameters for the bootstrapping method used to estimate the
sampling uncertainty in the evaluation of the predictions.
Three parameters are mandatory ('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), and one parameter
is optional ('seed'). If not provided, no bootstrapping is
performed. If provided, *dts* must also be provided.
dts: `List[str]`, 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"). If provided, it is only used if
*bootstrap* is also provided.
shape: (time,)
:Returns:
`List[numpy.ndarray]`
The sequence of evaluation metrics computed
in the same order as given in *metrics*.
shape: [(1, subsets, samples), ...]
)pbdoc",
"evald",
&evald_1d,
"Function to evaluate deterministic streamflow predictions (1D)",
py::arg("q_obs"), py::arg("q_prd"), py::arg("metrics"),
py::arg("transform") = "none",
py::arg("exponent") = 1,
......@@ -158,100 +66,11 @@ PYBIND11_MODULE(evalhyd, m)
py::dict("n_samples"_a=-9, "len_sample"_a=-9, "summary"_a=0),
py::arg("dts") = py::list()
);
m.def(
"evald", evalhyd::evald,
R"pbdoc(
Function to evaluate deterministic streamflow predictions.
:Parameters:
q_obs: `numpy.ndarray`
2D array of streamflow observations. Time steps with
missing observations must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (1, time)
q_prd: `numpy.ndarray`
2D array of streamflow predictions. Time steps with
missing predictions must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (series, time)
metrics: `List[str]`
The sequence of evaluation metrics to be computed.
transform: `str`, optional
The transformation to apply to both streamflow observations
and predictions prior to the calculation of the *metrics*.
exponent: `float`, optional
The value of the exponent n to use when the *transform* is
the power function. If not provided (or set to default value
1), the streamflow observations and predictions remain
untransformed.
epsilon: `float`, optional
The value of the small constant ε to add to both the
streamflow observations and predictions prior to the
calculation of the *metrics* when the *transform* is the
reciprocal function, the natural logarithm, or the power
function with a negative exponent (since none are defined
for 0). If not provided (or set to default value -9),
one hundredth of the mean of the streamflow observations
is used as value for epsilon.
t_msk: `numpy.ndarray`, optional
2D array of 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. If
provided, masks must feature the same number of dimensions as
observations and predictions, and it must broadcastable with
both of them.
shape: (subsets, time)
m_cdt: `numpy.ndarray`, optional
1D array of masking condition(s) to use to generate
temporal subsets. Each condition consists in a string and
can be specified on observed 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. If provided, only one condition per time series
of observations can be provided.
shape: (subsets,)
bootstrap: `dict`, optional
Parameters for the bootstrapping method used to estimate the
sampling uncertainty in the evaluation of the predictions.
Three parameters are mandatory ('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), and one parameter
is optional ('seed'). If not provided, no bootstrapping is
performed. If provided, *dts* must also be provided.
dts: `List[str]`, 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"). If provided, it is only used if
*bootstrap* is also provided.
shape: (time,)
:Returns:
`List[numpy.ndarray]`
The sequence of evaluation metrics computed
in the same order as given in *metrics*.
shape: [(series, subsets, samples), ...]
)pbdoc",
m.def(
"evald",
&evalhyd::evald<xt::pytensor<double, 2>, xt::pytensor<bool, 2>>,
"Function to evaluate deterministic streamflow predictions (2D)",
py::arg("q_obs"), py::arg("q_prd"), py::arg("metrics"),
py::arg("transform") = "none",
py::arg("exponent") = 1,
......@@ -265,86 +84,9 @@ PYBIND11_MODULE(evalhyd, m)
// probabilistic evaluation
m.def(
"evalp", evalhyd::evalp,
R"pbdoc(
Function to evaluate probabilistic streamflow predictions.
:Parameters:
q_obs: `numpy.ndarray`
2D array of streamflow observations. Time steps with
missing observations must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (sites, time)
q_prd: `numpy.ndarray`
4D array of streamflow predictions. Time steps with
missing predictions must be assigned `numpy.nan`
values. Those time steps will be ignored both in
the observations and in the predictions before the
*metrics* are computed.
shape: (sites, lead times, members, time)
metrics: `List[str]`
The sequence of evaluation metrics to be computed.
q_thr: `List[float]`, optional
The streamflow threshold(s) to consider for the *metrics*
assessing the prediction of exceedance events. If not
provided, set to default value as an empty `list`.
shape: (thresholds,)
t_msk: `numpy.ndarray`, optional
4D array of masks 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, 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)
m_cdt: `numpy.ndarray`, optional
2D array of conditions 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. If provided, as many sets of metrics are
returned as they are conditions provided.
shape: (sites, subsets)
bootstrap: `dict`, optional
Parameters for the bootstrapping method used to estimate the
sampling uncertainty in the evaluation of the predictions.
Three parameters are mandatory ('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), and one parameter
is optional ('seed'). If not provided, no bootstrapping is
performed. If provided, *dts* must also be provided.
dts: `List[str]`, 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"). If provided, it is only used if
*bootstrap* is also provided.
shape: (time,)
:Returns:
`List[numpy.ndarray]`
The sequence of evaluation metrics computed
in the same order as given in *metrics*.
shape: [(sites, lead times, subsets, samples, {quantiles,} {thresholds,} {components}), ...]
)pbdoc",
"evalp",
&evalhyd::evalp<xt::pytensor<double, 2>, xt::pytensor<double, 4>, xt::pytensor<bool, 4>>,
"Function to evaluate probabilistic streamflow predictions",
py::arg("q_obs"), py::arg("q_prd"), py::arg("metrics"),
py::arg("q_thr") = xt::pytensor<double, 2>({0}),
py::arg("t_msk") = xt::pytensor<bool, 4>({0}),
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment