Files
Time-Series-Analysis/venv/lib/python3.11/site-packages/prophet/stan_model/prophet.hpp
2025-08-01 04:33:03 -04:00

1343 lines
62 KiB
C++

// Code generated by stanc v2.33.1
#include <stan/model/model_header.hpp>
namespace prophet_model_namespace {
using stan::model::model_base_crtp;
using namespace stan::math;
stan::math::profile_map profiles__;
static constexpr std::array<const char*, 96> locations_array__ =
{" (found before start of program)",
" (in 'prophet.stan', line 110, column 2 to column 9)",
" (in 'prophet.stan', line 111, column 2 to column 9)",
" (in 'prophet.stan', line 112, column 2 to column 18)",
" (in 'prophet.stan', line 113, column 2 to column 26)",
" (in 'prophet.stan', line 114, column 2 to column 17)",
" (in 'prophet.stan', line 118, column 2 to column 18)",
" (in 'prophet.stan', line 124, column 4 to column 29)",
" (in 'prophet.stan', line 123, column 35 to line 125, column 3)",
" (in 'prophet.stan', line 123, column 9 to line 125, column 3)",
" (in 'prophet.stan', line 122, column 4 to column 64)",
" (in 'prophet.stan', line 121, column 35 to line 123, column 3)",
" (in 'prophet.stan', line 121, column 9 to line 125, column 3)",
" (in 'prophet.stan', line 120, column 4 to column 54)",
" (in 'prophet.stan', line 119, column 28 to line 121, column 3)",
" (in 'prophet.stan', line 119, column 2 to line 125, column 3)",
" (in 'prophet.stan', line 130, column 2 to column 19)",
" (in 'prophet.stan', line 131, column 2 to column 19)",
" (in 'prophet.stan', line 132, column 2 to column 37)",
" (in 'prophet.stan', line 133, column 2 to column 29)",
" (in 'prophet.stan', line 134, column 2 to column 27)",
" (in 'prophet.stan', line 137, column 2 to line 142, column 4)",
" (in 'prophet.stan', line 88, column 2 to column 8)",
" (in 'prophet.stan', line 89, column 2 to column 17)",
" (in 'prophet.stan', line 90, column 9 to column 10)",
" (in 'prophet.stan', line 90, column 2 to column 14)",
" (in 'prophet.stan', line 91, column 9 to column 10)",
" (in 'prophet.stan', line 91, column 2 to column 16)",
" (in 'prophet.stan', line 92, column 9 to column 10)",
" (in 'prophet.stan', line 92, column 2 to column 14)",
" (in 'prophet.stan', line 93, column 2 to column 8)",
" (in 'prophet.stan', line 94, column 9 to column 10)",
" (in 'prophet.stan', line 94, column 2 to column 21)",
" (in 'prophet.stan', line 95, column 9 to column 10)",
" (in 'prophet.stan', line 95, column 11 to column 12)",
" (in 'prophet.stan', line 95, column 2 to column 16)",
" (in 'prophet.stan', line 96, column 9 to column 10)",
" (in 'prophet.stan', line 96, column 2 to column 19)",
" (in 'prophet.stan', line 97, column 2 to column 20)",
" (in 'prophet.stan', line 98, column 2 to column 22)",
" (in 'prophet.stan', line 99, column 9 to column 10)",
" (in 'prophet.stan', line 99, column 2 to column 16)",
" (in 'prophet.stan', line 100, column 9 to column 10)",
" (in 'prophet.stan', line 100, column 2 to column 16)",
" (in 'prophet.stan', line 104, column 9 to column 10)",
" (in 'prophet.stan', line 104, column 12 to column 13)",
" (in 'prophet.stan', line 104, column 2 to column 61)",
" (in 'prophet.stan', line 105, column 9 to column 10)",
" (in 'prophet.stan', line 105, column 12 to column 13)",
" (in 'prophet.stan', line 105, column 2 to column 47)",
" (in 'prophet.stan', line 106, column 9 to column 10)",
" (in 'prophet.stan', line 106, column 12 to column 13)",
" (in 'prophet.stan', line 106, column 2 to column 47)",
" (in 'prophet.stan', line 112, column 9 to column 10)",
" (in 'prophet.stan', line 114, column 9 to column 10)",
" (in 'prophet.stan', line 118, column 9 to column 10)",
" (in 'prophet.stan', line 9, column 11 to column 12)",
" (in 'prophet.stan', line 9, column 14 to column 15)",
" (in 'prophet.stan', line 9, column 4 to column 19)",
" (in 'prophet.stan', line 10, column 15 to column 16)",
" (in 'prophet.stan', line 10, column 4 to column 24)",
" (in 'prophet.stan', line 11, column 4 to column 15)",
" (in 'prophet.stan', line 14, column 4 to column 28)",
" (in 'prophet.stan', line 15, column 4 to column 33)",
" (in 'prophet.stan', line 16, column 4 to column 15)",
" (in 'prophet.stan', line 21, column 8 to column 26)",
" (in 'prophet.stan', line 22, column 8 to column 28)",
" (in 'prophet.stan', line 20, column 58 to line 23, column 7)",
" (in 'prophet.stan', line 20, column 6 to line 23, column 7)",
" (in 'prophet.stan', line 24, column 6 to column 19)",
" (in 'prophet.stan', line 19, column 19 to line 25, column 5)",
" (in 'prophet.stan', line 19, column 4 to line 25, column 5)",
" (in 'prophet.stan', line 26, column 4 to column 13)",
" (in 'prophet.stan', line 7, column 73 to line 27, column 3)",
" (in 'prophet.stan', line 32, column 11 to column 12)",
" (in 'prophet.stan', line 32, column 4 to column 20)",
" (in 'prophet.stan', line 33, column 11 to column 16)",
" (in 'prophet.stan', line 33, column 4 to column 22)",
" (in 'prophet.stan', line 34, column 4 to column 14)",
" (in 'prophet.stan', line 37, column 4 to column 51)",
" (in 'prophet.stan', line 40, column 4 to column 13)",
" (in 'prophet.stan', line 42, column 6 to column 66)",
" (in 'prophet.stan', line 43, column 6 to column 29)",
" (in 'prophet.stan', line 41, column 19 to line 44, column 5)",
" (in 'prophet.stan', line 41, column 4 to line 44, column 5)",
" (in 'prophet.stan', line 45, column 4 to column 17)",
" (in 'prophet.stan', line 31, column 78 to line 46, column 3)",
" (in 'prophet.stan', line 58, column 11 to column 12)",
" (in 'prophet.stan', line 58, column 4 to column 20)",
" (in 'prophet.stan', line 60, column 4 to column 53)",
" (in 'prophet.stan', line 61, column 4 to column 70)",
" (in 'prophet.stan', line 57, column 4 to line 62, column 3)",
" (in 'prophet.stan', line 74, column 4 to column 65)",
" (in 'prophet.stan', line 73, column 4 to line 75, column 3)",
" (in 'prophet.stan', line 83, column 4 to column 28)",
" (in 'prophet.stan', line 82, column 4 to line 84, column 3)"};
template <typename T0__, typename T1__, typename T2__, typename T3__,
stan::require_all_t<stan::is_col_vector<T0__>,
stan::is_vt_not_complex<T0__>,
stan::is_col_vector<T1__>,
stan::is_vt_not_complex<T1__>,
std::is_integral<T2__>, std::is_integral<T3__>>* = nullptr>
Eigen::Matrix<stan::promote_args_t<stan::base_type_t<T0__>,
stan::base_type_t<T1__>>,-1,-1>
get_changepoint_matrix(const T0__& t_arg__, const T1__& t_change_arg__,
const T2__& T, const T3__& S, std::ostream* pstream__);
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
std::is_integral<T4__>>* = nullptr>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>>,-1,1>
logistic_gamma(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_change_arg__, const T4__& S, std::ostream*
pstream__);
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__, typename T5__, typename T6__, typename T7__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
stan::is_col_vector<T4__>,
stan::is_vt_not_complex<T4__>,
stan::is_eigen_matrix_dynamic<T5__>,
stan::is_vt_not_complex<T5__>,
stan::is_col_vector<T6__>,
stan::is_vt_not_complex<T6__>,
std::is_integral<T7__>>* = nullptr>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>, stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>,
stan::base_type_t<T6__>>>,-1,1>
logistic_trend(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_arg__, const T4__& cap_arg__, const T5__&
A_arg__, const T6__& t_change_arg__, const T7__& S,
std::ostream* pstream__);
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__, typename T5__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
stan::is_eigen_matrix_dynamic<T4__>,
stan::is_vt_not_complex<T4__>,
stan::is_col_vector<T5__>,
stan::is_vt_not_complex<T5__>>* = nullptr>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>, stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>>>,-1,1>
linear_trend(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_arg__, const T4__& A_arg__, const T5__&
t_change_arg__, std::ostream* pstream__);
template <typename T0__, typename T1__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
std::is_integral<T1__>>* = nullptr>
Eigen::Matrix<stan::promote_args_t<T0__>,-1,1>
flat_trend(const T0__& m, const T1__& T, std::ostream* pstream__);
// matrix get_changepoint_matrix(vector, vector, int, int)
template <typename T0__, typename T1__, typename T2__, typename T3__,
stan::require_all_t<stan::is_col_vector<T0__>,
stan::is_vt_not_complex<T0__>,
stan::is_col_vector<T1__>,
stan::is_vt_not_complex<T1__>,
std::is_integral<T2__>, std::is_integral<T3__>>*>
Eigen::Matrix<stan::promote_args_t<stan::base_type_t<T0__>,
stan::base_type_t<T1__>>,-1,-1>
get_changepoint_matrix(const T0__& t_arg__, const T1__& t_change_arg__,
const T2__& T, const T3__& S, std::ostream* pstream__) {
using local_scalar_t__ = stan::promote_args_t<stan::base_type_t<T0__>,
stan::base_type_t<T1__>>;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
const auto& t = stan::math::to_ref(t_arg__);
const auto& t_change = stan::math::to_ref(t_change_arg__);
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 56;
stan::math::validate_non_negative_index("A", "T", T);
current_statement__ = 57;
stan::math::validate_non_negative_index("A", "S", S);
Eigen::Matrix<local_scalar_t__,-1,-1> A =
Eigen::Matrix<local_scalar_t__,-1,-1>::Constant(T, S, DUMMY_VAR__);
current_statement__ = 59;
stan::math::validate_non_negative_index("a_row", "S", S);
Eigen::Matrix<local_scalar_t__,1,-1> a_row =
Eigen::Matrix<local_scalar_t__,1,-1>::Constant(S, DUMMY_VAR__);
int cp_idx = std::numeric_limits<int>::min();
current_statement__ = 62;
stan::model::assign(A, stan::math::rep_matrix(0, T, S),
"assigning variable A");
current_statement__ = 63;
stan::model::assign(a_row, stan::math::rep_row_vector(0, S),
"assigning variable a_row");
current_statement__ = 64;
cp_idx = 1;
current_statement__ = 71;
for (int i = 1; i <= T; ++i) {
current_statement__ = 68;
while ((stan::math::primitive_value(stan::math::logical_lte(cp_idx, S))
&&
stan::math::primitive_value(
stan::math::logical_gte(
stan::model::rvalue(t, "t", stan::model::index_uni(i)),
stan::model::rvalue(t_change, "t_change",
stan::model::index_uni(cp_idx)))))) {
current_statement__ = 65;
stan::model::assign(a_row, 1, "assigning variable a_row",
stan::model::index_uni(cp_idx));
current_statement__ = 66;
cp_idx = (cp_idx + 1);
}
current_statement__ = 69;
stan::model::assign(A, a_row, "assigning variable A",
stan::model::index_uni(i));
}
current_statement__ = 72;
return A;
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
// vector logistic_gamma(real, real, vector, vector, int)
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
std::is_integral<T4__>>*>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>>,-1,1>
logistic_gamma(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_change_arg__, const T4__& S, std::ostream*
pstream__) {
using local_scalar_t__ = stan::promote_args_t<T0__, T1__,
stan::base_type_t<T2__>,
stan::base_type_t<T3__>>;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
const auto& delta = stan::math::to_ref(delta_arg__);
const auto& t_change = stan::math::to_ref(t_change_arg__);
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 74;
stan::math::validate_non_negative_index("gamma", "S", S);
Eigen::Matrix<local_scalar_t__,-1,1> gamma =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
current_statement__ = 76;
stan::math::validate_non_negative_index("k_s", "S + 1", (S + 1));
Eigen::Matrix<local_scalar_t__,-1,1> k_s =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant((S + 1), DUMMY_VAR__);
local_scalar_t__ m_pr = DUMMY_VAR__;
current_statement__ = 79;
stan::model::assign(k_s,
stan::math::append_row(k,
stan::math::add(k, stan::math::cumulative_sum(delta))),
"assigning variable k_s");
current_statement__ = 80;
m_pr = m;
current_statement__ = 84;
for (int i = 1; i <= S; ++i) {
current_statement__ = 81;
stan::model::assign(gamma,
((stan::model::rvalue(t_change, "t_change", stan::model::index_uni(i))
- m_pr) * (1 -
(stan::model::rvalue(k_s, "k_s", stan::model::index_uni(i)) /
stan::model::rvalue(k_s, "k_s", stan::model::index_uni((i + 1)))))),
"assigning variable gamma", stan::model::index_uni(i));
current_statement__ = 82;
m_pr = (m_pr +
stan::model::rvalue(gamma, "gamma", stan::model::index_uni(i)));
}
current_statement__ = 85;
return gamma;
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
/* vector
logistic_trend(real, real, vector, vector, vector, matrix, vector, int)
*/
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__, typename T5__, typename T6__, typename T7__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
stan::is_col_vector<T4__>,
stan::is_vt_not_complex<T4__>,
stan::is_eigen_matrix_dynamic<T5__>,
stan::is_vt_not_complex<T5__>,
stan::is_col_vector<T6__>,
stan::is_vt_not_complex<T6__>,
std::is_integral<T7__>>*>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>, stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>,
stan::base_type_t<T6__>>>,-1,1>
logistic_trend(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_arg__, const T4__& cap_arg__, const T5__&
A_arg__, const T6__& t_change_arg__, const T7__& S,
std::ostream* pstream__) {
using local_scalar_t__ = stan::promote_args_t<T0__, T1__,
stan::base_type_t<T2__>,
stan::base_type_t<T3__>,
stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>,
stan::base_type_t<T6__>>>;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
const auto& delta = stan::math::to_ref(delta_arg__);
const auto& t = stan::math::to_ref(t_arg__);
const auto& cap = stan::math::to_ref(cap_arg__);
const auto& A = stan::math::to_ref(A_arg__);
const auto& t_change = stan::math::to_ref(t_change_arg__);
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 87;
stan::math::validate_non_negative_index("gamma", "S", S);
Eigen::Matrix<local_scalar_t__,-1,1> gamma =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
current_statement__ = 89;
stan::model::assign(gamma,
logistic_gamma(k, m, delta, t_change, S, pstream__),
"assigning variable gamma");
current_statement__ = 90;
return stan::math::elt_multiply(cap,
stan::math::inv_logit(
stan::math::elt_multiply(
stan::math::add(k, stan::math::multiply(A, delta)),
stan::math::subtract(t,
stan::math::add(m, stan::math::multiply(A, gamma))))));
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
// vector linear_trend(real, real, vector, vector, matrix, vector)
template <typename T0__, typename T1__, typename T2__, typename T3__,
typename T4__, typename T5__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
stan::math::disjunction<stan::is_autodiff<T1__>,
std::is_floating_point<T1__>>,
stan::is_col_vector<T2__>,
stan::is_vt_not_complex<T2__>,
stan::is_col_vector<T3__>,
stan::is_vt_not_complex<T3__>,
stan::is_eigen_matrix_dynamic<T4__>,
stan::is_vt_not_complex<T4__>,
stan::is_col_vector<T5__>,
stan::is_vt_not_complex<T5__>>*>
Eigen::Matrix<stan::promote_args_t<T0__, T1__, stan::base_type_t<T2__>,
stan::base_type_t<T3__>, stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>>>,-1,1>
linear_trend(const T0__& k, const T1__& m, const T2__& delta_arg__,
const T3__& t_arg__, const T4__& A_arg__, const T5__&
t_change_arg__, std::ostream* pstream__) {
using local_scalar_t__ = stan::promote_args_t<T0__, T1__,
stan::base_type_t<T2__>,
stan::base_type_t<T3__>,
stan::base_type_t<T4__>,
stan::promote_args_t<stan::base_type_t<T5__>>>;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
const auto& delta = stan::math::to_ref(delta_arg__);
const auto& t = stan::math::to_ref(t_arg__);
const auto& A = stan::math::to_ref(A_arg__);
const auto& t_change = stan::math::to_ref(t_change_arg__);
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 92;
return stan::math::add(
stan::math::elt_multiply(
stan::math::add(k, stan::math::multiply(A, delta)), t),
stan::math::add(m,
stan::math::multiply(A,
stan::math::elt_multiply(stan::math::minus(t_change), delta))));
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
// vector flat_trend(real, int)
template <typename T0__, typename T1__,
stan::require_all_t<stan::math::disjunction<stan::is_autodiff<T0__>,
std::is_floating_point<T0__>>,
std::is_integral<T1__>>*>
Eigen::Matrix<stan::promote_args_t<T0__>,-1,1>
flat_trend(const T0__& m, const T1__& T, std::ostream* pstream__) {
using local_scalar_t__ = stan::promote_args_t<T0__>;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 94;
return stan::math::rep_vector(m, T);
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
class prophet_model final : public model_base_crtp<prophet_model> {
private:
int T;
int K;
Eigen::Matrix<double,-1,1> t_data__;
Eigen::Matrix<double,-1,1> cap_data__;
Eigen::Matrix<double,-1,1> y_data__;
int S;
Eigen::Matrix<double,-1,1> t_change_data__;
Eigen::Matrix<double,-1,-1> X_data__;
Eigen::Matrix<double,-1,1> sigmas_data__;
double tau;
int trend_indicator;
Eigen::Matrix<double,-1,1> s_a_data__;
Eigen::Matrix<double,-1,1> s_m_data__;
Eigen::Matrix<double,-1,-1> A_data__;
Eigen::Matrix<double,-1,-1> X_sa_data__;
Eigen::Matrix<double,-1,-1> X_sm_data__;
Eigen::Map<Eigen::Matrix<double,-1,1>> t{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> cap{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> y{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> t_change{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,-1>> X{nullptr, 0, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> sigmas{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> s_a{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,1>> s_m{nullptr, 0};
Eigen::Map<Eigen::Matrix<double,-1,-1>> A{nullptr, 0, 0};
Eigen::Map<Eigen::Matrix<double,-1,-1>> X_sa{nullptr, 0, 0};
Eigen::Map<Eigen::Matrix<double,-1,-1>> X_sm{nullptr, 0, 0};
public:
~prophet_model() {}
prophet_model(stan::io::var_context& context__, unsigned int
random_seed__ = 0, std::ostream* pstream__ = nullptr)
: model_base_crtp(0) {
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
using local_scalar_t__ = double;
boost::ecuyer1988 base_rng__ =
stan::services::util::create_rng(random_seed__, 0);
// suppress unused var warning
(void) base_rng__;
static constexpr const char* function__ =
"prophet_model_namespace::prophet_model";
// suppress unused var warning
(void) function__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
int pos__ = std::numeric_limits<int>::min();
pos__ = 1;
current_statement__ = 22;
context__.validate_dims("data initialization", "T", "int",
std::vector<size_t>{});
T = std::numeric_limits<int>::min();
current_statement__ = 22;
T = context__.vals_i("T")[(1 - 1)];
current_statement__ = 23;
context__.validate_dims("data initialization", "K", "int",
std::vector<size_t>{});
K = std::numeric_limits<int>::min();
current_statement__ = 23;
K = context__.vals_i("K")[(1 - 1)];
current_statement__ = 23;
stan::math::check_greater_or_equal(function__, "K", K, 1);
current_statement__ = 24;
stan::math::validate_non_negative_index("t", "T", T);
current_statement__ = 25;
context__.validate_dims("data initialization", "t", "double",
std::vector<size_t>{static_cast<size_t>(T)});
t_data__ = Eigen::Matrix<double,-1,1>::Constant(T,
std::numeric_limits<double>::quiet_NaN());
new (&t) Eigen::Map<Eigen::Matrix<double,-1,1>>(t_data__.data(), T);
{
std::vector<local_scalar_t__> t_flat__;
current_statement__ = 25;
t_flat__ = context__.vals_r("t");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= T; ++sym1__) {
stan::model::assign(t, t_flat__[(pos__ - 1)],
"assigning variable t", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 26;
stan::math::validate_non_negative_index("cap", "T", T);
current_statement__ = 27;
context__.validate_dims("data initialization", "cap", "double",
std::vector<size_t>{static_cast<size_t>(T)});
cap_data__ = Eigen::Matrix<double,-1,1>::Constant(T,
std::numeric_limits<double>::quiet_NaN());
new (&cap) Eigen::Map<Eigen::Matrix<double,-1,1>>(cap_data__.data(), T);
{
std::vector<local_scalar_t__> cap_flat__;
current_statement__ = 27;
cap_flat__ = context__.vals_r("cap");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= T; ++sym1__) {
stan::model::assign(cap, cap_flat__[(pos__ - 1)],
"assigning variable cap", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 28;
stan::math::validate_non_negative_index("y", "T", T);
current_statement__ = 29;
context__.validate_dims("data initialization", "y", "double",
std::vector<size_t>{static_cast<size_t>(T)});
y_data__ = Eigen::Matrix<double,-1,1>::Constant(T,
std::numeric_limits<double>::quiet_NaN());
new (&y) Eigen::Map<Eigen::Matrix<double,-1,1>>(y_data__.data(), T);
{
std::vector<local_scalar_t__> y_flat__;
current_statement__ = 29;
y_flat__ = context__.vals_r("y");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= T; ++sym1__) {
stan::model::assign(y, y_flat__[(pos__ - 1)],
"assigning variable y", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 30;
context__.validate_dims("data initialization", "S", "int",
std::vector<size_t>{});
S = std::numeric_limits<int>::min();
current_statement__ = 30;
S = context__.vals_i("S")[(1 - 1)];
current_statement__ = 31;
stan::math::validate_non_negative_index("t_change", "S", S);
current_statement__ = 32;
context__.validate_dims("data initialization", "t_change", "double",
std::vector<size_t>{static_cast<size_t>(S)});
t_change_data__ = Eigen::Matrix<double,-1,1>::Constant(S,
std::numeric_limits<double>::quiet_NaN());
new (&t_change)
Eigen::Map<Eigen::Matrix<double,-1,1>>(t_change_data__.data(), S);
{
std::vector<local_scalar_t__> t_change_flat__;
current_statement__ = 32;
t_change_flat__ = context__.vals_r("t_change");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= S; ++sym1__) {
stan::model::assign(t_change, t_change_flat__[(pos__ - 1)],
"assigning variable t_change", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 33;
stan::math::validate_non_negative_index("X", "T", T);
current_statement__ = 34;
stan::math::validate_non_negative_index("X", "K", K);
current_statement__ = 35;
context__.validate_dims("data initialization", "X", "double",
std::vector<size_t>{static_cast<size_t>(T), static_cast<size_t>(K)});
X_data__ = Eigen::Matrix<double,-1,-1>::Constant(T, K,
std::numeric_limits<double>::quiet_NaN());
new (&X) Eigen::Map<Eigen::Matrix<double,-1,-1>>(X_data__.data(), T, K);
{
std::vector<local_scalar_t__> X_flat__;
current_statement__ = 35;
X_flat__ = context__.vals_r("X");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
for (int sym2__ = 1; sym2__ <= T; ++sym2__) {
stan::model::assign(X, X_flat__[(pos__ - 1)],
"assigning variable X", stan::model::index_uni(sym2__),
stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
}
current_statement__ = 36;
stan::math::validate_non_negative_index("sigmas", "K", K);
current_statement__ = 37;
context__.validate_dims("data initialization", "sigmas", "double",
std::vector<size_t>{static_cast<size_t>(K)});
sigmas_data__ = Eigen::Matrix<double,-1,1>::Constant(K,
std::numeric_limits<double>::quiet_NaN());
new (&sigmas)
Eigen::Map<Eigen::Matrix<double,-1,1>>(sigmas_data__.data(), K);
{
std::vector<local_scalar_t__> sigmas_flat__;
current_statement__ = 37;
sigmas_flat__ = context__.vals_r("sigmas");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
stan::model::assign(sigmas, sigmas_flat__[(pos__ - 1)],
"assigning variable sigmas", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 38;
context__.validate_dims("data initialization", "tau", "double",
std::vector<size_t>{});
tau = std::numeric_limits<double>::quiet_NaN();
current_statement__ = 38;
tau = context__.vals_r("tau")[(1 - 1)];
current_statement__ = 38;
stan::math::check_greater_or_equal(function__, "tau", tau, 0);
current_statement__ = 39;
context__.validate_dims("data initialization", "trend_indicator",
"int", std::vector<size_t>{});
trend_indicator = std::numeric_limits<int>::min();
current_statement__ = 39;
trend_indicator = context__.vals_i("trend_indicator")[(1 - 1)];
current_statement__ = 40;
stan::math::validate_non_negative_index("s_a", "K", K);
current_statement__ = 41;
context__.validate_dims("data initialization", "s_a", "double",
std::vector<size_t>{static_cast<size_t>(K)});
s_a_data__ = Eigen::Matrix<double,-1,1>::Constant(K,
std::numeric_limits<double>::quiet_NaN());
new (&s_a) Eigen::Map<Eigen::Matrix<double,-1,1>>(s_a_data__.data(), K);
{
std::vector<local_scalar_t__> s_a_flat__;
current_statement__ = 41;
s_a_flat__ = context__.vals_r("s_a");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
stan::model::assign(s_a, s_a_flat__[(pos__ - 1)],
"assigning variable s_a", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 42;
stan::math::validate_non_negative_index("s_m", "K", K);
current_statement__ = 43;
context__.validate_dims("data initialization", "s_m", "double",
std::vector<size_t>{static_cast<size_t>(K)});
s_m_data__ = Eigen::Matrix<double,-1,1>::Constant(K,
std::numeric_limits<double>::quiet_NaN());
new (&s_m) Eigen::Map<Eigen::Matrix<double,-1,1>>(s_m_data__.data(), K);
{
std::vector<local_scalar_t__> s_m_flat__;
current_statement__ = 43;
s_m_flat__ = context__.vals_r("s_m");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
stan::model::assign(s_m, s_m_flat__[(pos__ - 1)],
"assigning variable s_m", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
current_statement__ = 44;
stan::math::validate_non_negative_index("A", "T", T);
current_statement__ = 45;
stan::math::validate_non_negative_index("A", "S", S);
current_statement__ = 46;
A_data__ = Eigen::Matrix<double,-1,-1>::Constant(T, S,
std::numeric_limits<double>::quiet_NaN());
new (&A) Eigen::Map<Eigen::Matrix<double,-1,-1>>(A_data__.data(), T, S);
current_statement__ = 46;
stan::model::assign(A,
get_changepoint_matrix(t, t_change, T, S, pstream__),
"assigning variable A");
current_statement__ = 47;
stan::math::validate_non_negative_index("X_sa", "T", T);
current_statement__ = 48;
stan::math::validate_non_negative_index("X_sa", "K", K);
current_statement__ = 49;
X_sa_data__ = Eigen::Matrix<double,-1,-1>::Constant(T, K,
std::numeric_limits<double>::quiet_NaN());
new (&X_sa) Eigen::Map<Eigen::Matrix<double,-1,-1>>(X_sa_data__.data(),
T, K);
current_statement__ = 49;
stan::model::assign(X_sa,
stan::math::elt_multiply(X,
stan::math::rep_matrix(stan::math::transpose(s_a), T)),
"assigning variable X_sa");
current_statement__ = 50;
stan::math::validate_non_negative_index("X_sm", "T", T);
current_statement__ = 51;
stan::math::validate_non_negative_index("X_sm", "K", K);
current_statement__ = 52;
X_sm_data__ = Eigen::Matrix<double,-1,-1>::Constant(T, K,
std::numeric_limits<double>::quiet_NaN());
new (&X_sm) Eigen::Map<Eigen::Matrix<double,-1,-1>>(X_sm_data__.data(),
T, K);
current_statement__ = 52;
stan::model::assign(X_sm,
stan::math::elt_multiply(X,
stan::math::rep_matrix(stan::math::transpose(s_m), T)),
"assigning variable X_sm");
current_statement__ = 53;
stan::math::validate_non_negative_index("delta", "S", S);
current_statement__ = 54;
stan::math::validate_non_negative_index("beta", "K", K);
current_statement__ = 55;
stan::math::validate_non_negative_index("trend", "T", T);
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
num_params_r__ = 1 + 1 + S + 1 + K;
}
inline std::string model_name() const final {
return "prophet_model";
}
inline std::vector<std::string> model_compile_info() const noexcept {
return std::vector<std::string>{"stanc_version = stanc3 v2.33.1",
"stancflags = --filename-in-msg=prophet.stan"};
}
// Base log prob
template <bool propto__, bool jacobian__, typename VecR, typename VecI,
stan::require_vector_like_t<VecR>* = nullptr,
stan::require_vector_like_vt<std::is_integral, VecI>* = nullptr,
stan::require_not_st_var<VecR>* = nullptr>
inline stan::scalar_type_t<VecR>
log_prob_impl(VecR& params_r__, VecI& params_i__, std::ostream*
pstream__ = nullptr) const {
using T__ = stan::scalar_type_t<VecR>;
using local_scalar_t__ = T__;
T__ lp__(0.0);
stan::math::accumulator<T__> lp_accum__;
stan::io::deserializer<local_scalar_t__> in__(params_r__, params_i__);
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
static constexpr const char* function__ =
"prophet_model_namespace::log_prob";
// suppress unused var warning
(void) function__;
try {
local_scalar_t__ k = DUMMY_VAR__;
current_statement__ = 1;
k = in__.template read<local_scalar_t__>();
local_scalar_t__ m = DUMMY_VAR__;
current_statement__ = 2;
m = in__.template read<local_scalar_t__>();
Eigen::Matrix<local_scalar_t__,-1,1> delta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
current_statement__ = 3;
delta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(S);
local_scalar_t__ sigma_obs = DUMMY_VAR__;
current_statement__ = 4;
sigma_obs = in__.template read_constrain_lb<local_scalar_t__,
jacobian__>(0, lp__);
Eigen::Matrix<local_scalar_t__,-1,1> beta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(K, DUMMY_VAR__);
current_statement__ = 5;
beta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(K);
Eigen::Matrix<local_scalar_t__,-1,1> trend =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(T, DUMMY_VAR__);
current_statement__ = 15;
if (stan::math::logical_eq(trend_indicator, 0)) {
current_statement__ = 13;
stan::model::assign(trend,
linear_trend(k, m, delta, t, A, t_change, pstream__),
"assigning variable trend");
} else {
current_statement__ = 12;
if (stan::math::logical_eq(trend_indicator, 1)) {
current_statement__ = 10;
stan::model::assign(trend,
logistic_trend(k, m, delta, t, cap, A, t_change, S, pstream__),
"assigning variable trend");
} else {
current_statement__ = 9;
if (stan::math::logical_eq(trend_indicator, 2)) {
current_statement__ = 7;
stan::model::assign(trend, flat_trend(m, T, pstream__),
"assigning variable trend");
}
}
}
{
current_statement__ = 16;
lp_accum__.add(stan::math::normal_lpdf<propto__>(k, 0, 5));
current_statement__ = 17;
lp_accum__.add(stan::math::normal_lpdf<propto__>(m, 0, 5));
current_statement__ = 18;
lp_accum__.add(stan::math::double_exponential_lpdf<propto__>(delta,
0, tau));
current_statement__ = 19;
lp_accum__.add(stan::math::normal_lpdf<propto__>(sigma_obs, 0, 0.5));
current_statement__ = 20;
lp_accum__.add(stan::math::normal_lpdf<propto__>(beta, 0, sigmas));
current_statement__ = 21;
lp_accum__.add(stan::math::normal_id_glm_lpdf<propto__>(y, X_sa,
stan::math::elt_multiply(trend,
stan::math::add(1,
stan::math::multiply(X_sm, beta))), beta,
sigma_obs));
}
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
lp_accum__.add(lp__);
return lp_accum__.sum();
}
// Reverse mode autodiff log prob
template <bool propto__, bool jacobian__, typename VecR, typename VecI,
stan::require_vector_like_t<VecR>* = nullptr,
stan::require_vector_like_vt<std::is_integral, VecI>* = nullptr,
stan::require_st_var<VecR>* = nullptr>
inline stan::scalar_type_t<VecR>
log_prob_impl(VecR& params_r__, VecI& params_i__, std::ostream*
pstream__ = nullptr) const {
using T__ = stan::scalar_type_t<VecR>;
using local_scalar_t__ = T__;
T__ lp__(0.0);
stan::math::accumulator<T__> lp_accum__;
stan::io::deserializer<local_scalar_t__> in__(params_r__, params_i__);
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
static constexpr const char* function__ =
"prophet_model_namespace::log_prob";
// suppress unused var warning
(void) function__;
try {
local_scalar_t__ k = DUMMY_VAR__;
current_statement__ = 1;
k = in__.template read<local_scalar_t__>();
local_scalar_t__ m = DUMMY_VAR__;
current_statement__ = 2;
m = in__.template read<local_scalar_t__>();
Eigen::Matrix<local_scalar_t__,-1,1> delta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
current_statement__ = 3;
delta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(S);
local_scalar_t__ sigma_obs = DUMMY_VAR__;
current_statement__ = 4;
sigma_obs = in__.template read_constrain_lb<local_scalar_t__,
jacobian__>(0, lp__);
Eigen::Matrix<local_scalar_t__,-1,1> beta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(K, DUMMY_VAR__);
current_statement__ = 5;
beta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(K);
Eigen::Matrix<local_scalar_t__,-1,1> trend =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(T, DUMMY_VAR__);
current_statement__ = 15;
if (stan::math::logical_eq(trend_indicator, 0)) {
current_statement__ = 13;
stan::model::assign(trend,
linear_trend(k, m, delta, t, A, t_change, pstream__),
"assigning variable trend");
} else {
current_statement__ = 12;
if (stan::math::logical_eq(trend_indicator, 1)) {
current_statement__ = 10;
stan::model::assign(trend,
logistic_trend(k, m, delta, t, cap, A, t_change, S, pstream__),
"assigning variable trend");
} else {
current_statement__ = 9;
if (stan::math::logical_eq(trend_indicator, 2)) {
current_statement__ = 7;
stan::model::assign(trend, flat_trend(m, T, pstream__),
"assigning variable trend");
}
}
}
{
current_statement__ = 16;
lp_accum__.add(stan::math::normal_lpdf<propto__>(k, 0, 5));
current_statement__ = 17;
lp_accum__.add(stan::math::normal_lpdf<propto__>(m, 0, 5));
current_statement__ = 18;
lp_accum__.add(stan::math::double_exponential_lpdf<propto__>(delta,
0, tau));
current_statement__ = 19;
lp_accum__.add(stan::math::normal_lpdf<propto__>(sigma_obs, 0, 0.5));
current_statement__ = 20;
lp_accum__.add(stan::math::normal_lpdf<propto__>(beta, 0, sigmas));
current_statement__ = 21;
lp_accum__.add(stan::math::normal_id_glm_lpdf<propto__>(y, X_sa,
stan::math::elt_multiply(trend,
stan::math::add(1,
stan::math::multiply(X_sm, beta))), beta,
sigma_obs));
}
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
lp_accum__.add(lp__);
return lp_accum__.sum();
}
template <typename RNG, typename VecR, typename VecI, typename VecVar,
stan::require_vector_like_vt<std::is_floating_point,
VecR>* = nullptr, stan::require_vector_like_vt<std::is_integral,
VecI>* = nullptr, stan::require_vector_vt<std::is_floating_point,
VecVar>* = nullptr>
inline void
write_array_impl(RNG& base_rng__, VecR& params_r__, VecI& params_i__,
VecVar& vars__, const bool
emit_transformed_parameters__ = true, const bool
emit_generated_quantities__ = true, std::ostream*
pstream__ = nullptr) const {
using local_scalar_t__ = double;
stan::io::deserializer<local_scalar_t__> in__(params_r__, params_i__);
stan::io::serializer<local_scalar_t__> out__(vars__);
static constexpr bool propto__ = true;
// suppress unused var warning
(void) propto__;
double lp__ = 0.0;
// suppress unused var warning
(void) lp__;
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
stan::math::accumulator<double> lp_accum__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
constexpr bool jacobian__ = false;
// suppress unused var warning
(void) jacobian__;
static constexpr const char* function__ =
"prophet_model_namespace::write_array";
// suppress unused var warning
(void) function__;
try {
double k = std::numeric_limits<double>::quiet_NaN();
current_statement__ = 1;
k = in__.template read<local_scalar_t__>();
double m = std::numeric_limits<double>::quiet_NaN();
current_statement__ = 2;
m = in__.template read<local_scalar_t__>();
Eigen::Matrix<double,-1,1> delta =
Eigen::Matrix<double,-1,1>::Constant(S,
std::numeric_limits<double>::quiet_NaN());
current_statement__ = 3;
delta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(S);
double sigma_obs = std::numeric_limits<double>::quiet_NaN();
current_statement__ = 4;
sigma_obs = in__.template read_constrain_lb<local_scalar_t__,
jacobian__>(0, lp__);
Eigen::Matrix<double,-1,1> beta =
Eigen::Matrix<double,-1,1>::Constant(K,
std::numeric_limits<double>::quiet_NaN());
current_statement__ = 5;
beta = in__.template read<Eigen::Matrix<local_scalar_t__,-1,1>>(K);
Eigen::Matrix<double,-1,1> trend =
Eigen::Matrix<double,-1,1>::Constant(T,
std::numeric_limits<double>::quiet_NaN());
out__.write(k);
out__.write(m);
out__.write(delta);
out__.write(sigma_obs);
out__.write(beta);
if (stan::math::logical_negation(
(stan::math::primitive_value(emit_transformed_parameters__) ||
stan::math::primitive_value(emit_generated_quantities__)))) {
return ;
}
current_statement__ = 15;
if (stan::math::logical_eq(trend_indicator, 0)) {
current_statement__ = 13;
stan::model::assign(trend,
linear_trend(k, m, delta, t, A, t_change, pstream__),
"assigning variable trend");
} else {
current_statement__ = 12;
if (stan::math::logical_eq(trend_indicator, 1)) {
current_statement__ = 10;
stan::model::assign(trend,
logistic_trend(k, m, delta, t, cap, A, t_change, S, pstream__),
"assigning variable trend");
} else {
current_statement__ = 9;
if (stan::math::logical_eq(trend_indicator, 2)) {
current_statement__ = 7;
stan::model::assign(trend, flat_trend(m, T, pstream__),
"assigning variable trend");
}
}
}
if (emit_transformed_parameters__) {
out__.write(trend);
}
if (stan::math::logical_negation(emit_generated_quantities__)) {
return ;
}
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
template <typename VecVar, typename VecI,
stan::require_vector_t<VecVar>* = nullptr,
stan::require_vector_like_vt<std::is_integral, VecI>* = nullptr>
inline void
unconstrain_array_impl(const VecVar& params_r__, const VecI& params_i__,
VecVar& vars__, std::ostream* pstream__ = nullptr) const {
using local_scalar_t__ = double;
stan::io::deserializer<local_scalar_t__> in__(params_r__, params_i__);
stan::io::serializer<local_scalar_t__> out__(vars__);
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
local_scalar_t__ k = DUMMY_VAR__;
current_statement__ = 1;
k = in__.read<local_scalar_t__>();
out__.write(k);
local_scalar_t__ m = DUMMY_VAR__;
current_statement__ = 2;
m = in__.read<local_scalar_t__>();
out__.write(m);
Eigen::Matrix<local_scalar_t__,-1,1> delta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
current_statement__ = 3;
stan::model::assign(delta,
in__.read<Eigen::Matrix<local_scalar_t__,-1,1>>(S),
"assigning variable delta");
out__.write(delta);
local_scalar_t__ sigma_obs = DUMMY_VAR__;
current_statement__ = 4;
sigma_obs = in__.read<local_scalar_t__>();
out__.write_free_lb(0, sigma_obs);
Eigen::Matrix<local_scalar_t__,-1,1> beta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(K, DUMMY_VAR__);
current_statement__ = 5;
stan::model::assign(beta,
in__.read<Eigen::Matrix<local_scalar_t__,-1,1>>(K),
"assigning variable beta");
out__.write(beta);
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
template <typename VecVar, stan::require_vector_t<VecVar>* = nullptr>
inline void
transform_inits_impl(const stan::io::var_context& context__, VecVar&
vars__, std::ostream* pstream__ = nullptr) const {
using local_scalar_t__ = double;
stan::io::serializer<local_scalar_t__> out__(vars__);
int current_statement__ = 0;
// suppress unused var warning
(void) current_statement__;
local_scalar_t__ DUMMY_VAR__(std::numeric_limits<double>::quiet_NaN());
// suppress unused var warning
(void) DUMMY_VAR__;
try {
current_statement__ = 1;
context__.validate_dims("parameter initialization", "k", "double",
std::vector<size_t>{});
current_statement__ = 2;
context__.validate_dims("parameter initialization", "m", "double",
std::vector<size_t>{});
current_statement__ = 3;
context__.validate_dims("parameter initialization", "delta", "double",
std::vector<size_t>{static_cast<size_t>(S)});
current_statement__ = 4;
context__.validate_dims("parameter initialization", "sigma_obs",
"double", std::vector<size_t>{});
current_statement__ = 5;
context__.validate_dims("parameter initialization", "beta", "double",
std::vector<size_t>{static_cast<size_t>(K)});
int pos__ = std::numeric_limits<int>::min();
pos__ = 1;
local_scalar_t__ k = DUMMY_VAR__;
current_statement__ = 1;
k = context__.vals_r("k")[(1 - 1)];
out__.write(k);
local_scalar_t__ m = DUMMY_VAR__;
current_statement__ = 2;
m = context__.vals_r("m")[(1 - 1)];
out__.write(m);
Eigen::Matrix<local_scalar_t__,-1,1> delta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(S, DUMMY_VAR__);
{
std::vector<local_scalar_t__> delta_flat__;
current_statement__ = 3;
delta_flat__ = context__.vals_r("delta");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= S; ++sym1__) {
stan::model::assign(delta, delta_flat__[(pos__ - 1)],
"assigning variable delta", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
out__.write(delta);
local_scalar_t__ sigma_obs = DUMMY_VAR__;
current_statement__ = 4;
sigma_obs = context__.vals_r("sigma_obs")[(1 - 1)];
out__.write_free_lb(0, sigma_obs);
Eigen::Matrix<local_scalar_t__,-1,1> beta =
Eigen::Matrix<local_scalar_t__,-1,1>::Constant(K, DUMMY_VAR__);
{
std::vector<local_scalar_t__> beta_flat__;
current_statement__ = 5;
beta_flat__ = context__.vals_r("beta");
pos__ = 1;
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
stan::model::assign(beta, beta_flat__[(pos__ - 1)],
"assigning variable beta", stan::model::index_uni(sym1__));
pos__ = (pos__ + 1);
}
}
out__.write(beta);
} catch (const std::exception& e) {
stan::lang::rethrow_located(e, locations_array__[current_statement__]);
}
}
inline void
get_param_names(std::vector<std::string>& names__, const bool
emit_transformed_parameters__ = true, const bool
emit_generated_quantities__ = true) const {
names__ = std::vector<std::string>{"k", "m", "delta", "sigma_obs",
"beta"};
if (emit_transformed_parameters__) {
std::vector<std::string> temp{"trend"};
names__.reserve(names__.size() + temp.size());
names__.insert(names__.end(), temp.begin(), temp.end());
}
if (emit_generated_quantities__) {}
}
inline void
get_dims(std::vector<std::vector<size_t>>& dimss__, const bool
emit_transformed_parameters__ = true, const bool
emit_generated_quantities__ = true) const {
dimss__ = std::vector<std::vector<size_t>>{std::vector<size_t>{},
std::vector<size_t>{},
std::vector<size_t>{static_cast<size_t>(S)},
std::vector<size_t>{},
std::vector<size_t>{static_cast<size_t>(K)}};
if (emit_transformed_parameters__) {
std::vector<std::vector<size_t>>
temp{std::vector<size_t>{static_cast<size_t>(T)}};
dimss__.reserve(dimss__.size() + temp.size());
dimss__.insert(dimss__.end(), temp.begin(), temp.end());
}
if (emit_generated_quantities__) {}
}
inline void
constrained_param_names(std::vector<std::string>& param_names__, bool
emit_transformed_parameters__ = true, bool
emit_generated_quantities__ = true) const final {
param_names__.emplace_back(std::string() + "k");
param_names__.emplace_back(std::string() + "m");
for (int sym1__ = 1; sym1__ <= S; ++sym1__) {
param_names__.emplace_back(std::string() + "delta" + '.' +
std::to_string(sym1__));
}
param_names__.emplace_back(std::string() + "sigma_obs");
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
param_names__.emplace_back(std::string() + "beta" + '.' +
std::to_string(sym1__));
}
if (emit_transformed_parameters__) {
for (int sym1__ = 1; sym1__ <= T; ++sym1__) {
param_names__.emplace_back(std::string() + "trend" + '.' +
std::to_string(sym1__));
}
}
if (emit_generated_quantities__) {}
}
inline void
unconstrained_param_names(std::vector<std::string>& param_names__, bool
emit_transformed_parameters__ = true, bool
emit_generated_quantities__ = true) const final {
param_names__.emplace_back(std::string() + "k");
param_names__.emplace_back(std::string() + "m");
for (int sym1__ = 1; sym1__ <= S; ++sym1__) {
param_names__.emplace_back(std::string() + "delta" + '.' +
std::to_string(sym1__));
}
param_names__.emplace_back(std::string() + "sigma_obs");
for (int sym1__ = 1; sym1__ <= K; ++sym1__) {
param_names__.emplace_back(std::string() + "beta" + '.' +
std::to_string(sym1__));
}
if (emit_transformed_parameters__) {
for (int sym1__ = 1; sym1__ <= T; ++sym1__) {
param_names__.emplace_back(std::string() + "trend" + '.' +
std::to_string(sym1__));
}
}
if (emit_generated_quantities__) {}
}
inline std::string get_constrained_sizedtypes() const {
return std::string("[{\"name\":\"k\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"m\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"delta\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(S) + "},\"block\":\"parameters\"},{\"name\":\"sigma_obs\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"beta\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(K) + "},\"block\":\"parameters\"},{\"name\":\"trend\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(T) + "},\"block\":\"transformed_parameters\"}]");
}
inline std::string get_unconstrained_sizedtypes() const {
return std::string("[{\"name\":\"k\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"m\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"delta\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(S) + "},\"block\":\"parameters\"},{\"name\":\"sigma_obs\",\"type\":{\"name\":\"real\"},\"block\":\"parameters\"},{\"name\":\"beta\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(K) + "},\"block\":\"parameters\"},{\"name\":\"trend\",\"type\":{\"name\":\"vector\",\"length\":" + std::to_string(T) + "},\"block\":\"transformed_parameters\"}]");
}
// Begin method overload boilerplate
template <typename RNG> inline void
write_array(RNG& base_rng, Eigen::Matrix<double,-1,1>& params_r,
Eigen::Matrix<double,-1,1>& vars, const bool
emit_transformed_parameters = true, const bool
emit_generated_quantities = true, std::ostream*
pstream = nullptr) const {
const size_t num_params__ = ((((1 + 1) + S) + 1) + K);
const size_t num_transformed = emit_transformed_parameters * (T);
const size_t num_gen_quantities = emit_generated_quantities * (0);
const size_t num_to_write = num_params__ + num_transformed +
num_gen_quantities;
std::vector<int> params_i;
vars = Eigen::Matrix<double,-1,1>::Constant(num_to_write,
std::numeric_limits<double>::quiet_NaN());
write_array_impl(base_rng, params_r, params_i, vars,
emit_transformed_parameters, emit_generated_quantities, pstream);
}
template <typename RNG> inline void
write_array(RNG& base_rng, std::vector<double>& params_r, std::vector<int>&
params_i, std::vector<double>& vars, bool
emit_transformed_parameters = true, bool
emit_generated_quantities = true, std::ostream*
pstream = nullptr) const {
const size_t num_params__ = ((((1 + 1) + S) + 1) + K);
const size_t num_transformed = emit_transformed_parameters * (T);
const size_t num_gen_quantities = emit_generated_quantities * (0);
const size_t num_to_write = num_params__ + num_transformed +
num_gen_quantities;
vars = std::vector<double>(num_to_write,
std::numeric_limits<double>::quiet_NaN());
write_array_impl(base_rng, params_r, params_i, vars,
emit_transformed_parameters, emit_generated_quantities, pstream);
}
template <bool propto__, bool jacobian__, typename T_> inline T_
log_prob(Eigen::Matrix<T_,-1,1>& params_r, std::ostream* pstream = nullptr) const {
Eigen::Matrix<int,-1,1> params_i;
return log_prob_impl<propto__, jacobian__>(params_r, params_i, pstream);
}
template <bool propto__, bool jacobian__, typename T_> inline T_
log_prob(std::vector<T_>& params_r, std::vector<int>& params_i,
std::ostream* pstream = nullptr) const {
return log_prob_impl<propto__, jacobian__>(params_r, params_i, pstream);
}
inline void
transform_inits(const stan::io::var_context& context,
Eigen::Matrix<double,-1,1>& params_r, std::ostream*
pstream = nullptr) const final {
std::vector<double> params_r_vec(params_r.size());
std::vector<int> params_i;
transform_inits(context, params_i, params_r_vec, pstream);
params_r = Eigen::Map<Eigen::Matrix<double,-1,1>>(params_r_vec.data(),
params_r_vec.size());
}
inline void
transform_inits(const stan::io::var_context& context, std::vector<int>&
params_i, std::vector<double>& vars, std::ostream*
pstream__ = nullptr) const {
vars.resize(num_params_r__);
transform_inits_impl(context, vars, pstream__);
}
inline void
unconstrain_array(const std::vector<double>& params_constrained,
std::vector<double>& params_unconstrained, std::ostream*
pstream = nullptr) const {
const std::vector<int> params_i;
params_unconstrained = std::vector<double>(num_params_r__,
std::numeric_limits<double>::quiet_NaN());
unconstrain_array_impl(params_constrained, params_i,
params_unconstrained, pstream);
}
inline void
unconstrain_array(const Eigen::Matrix<double,-1,1>& params_constrained,
Eigen::Matrix<double,-1,1>& params_unconstrained,
std::ostream* pstream = nullptr) const {
const std::vector<int> params_i;
params_unconstrained = Eigen::Matrix<double,-1,1>::Constant(num_params_r__,
std::numeric_limits<double>::quiet_NaN());
unconstrain_array_impl(params_constrained, params_i,
params_unconstrained, pstream);
}
};
}
using stan_model = prophet_model_namespace::prophet_model;
#ifndef USING_R
// Boilerplate
stan::model::model_base&
new_model(stan::io::var_context& data_context, unsigned int seed,
std::ostream* msg_stream) {
stan_model* m = new stan_model(data_context, seed, msg_stream);
return *m;
}
stan::math::profile_map& get_stan_profile_data() {
return prophet_model_namespace::profiles__;
}
#endif