ExaGeoStat
ExaGeoStat is a parallel high performance unified framework for geostatistics on manycore systems.
|
Go to the source code of this file.
Functions | |
void | core_dcmg (double *A, int m, int n, int m0, int n0, location *l1, location *l2, double *localtheta, int distance_metric) |
double | core_dmdet (double *A, int m, int n, int m0, int n0) |
void | core_dzcpy (double *Z, int m, int m0, double *r) |
double | core_ddotp (double *Z, double *dotproduct, int n) |
Copyright (c) 2017-2018 King Abdullah University of Science and Technology All rights reserved.
ExaGeoStat is a software package provided by KAUST
Core functions header file.
Definition in file exageostatcore.h.
void core_dcmg | ( | double * | A, |
int | m, | ||
int | n, | ||
int | m0, | ||
int | n0, | ||
location * | l1, | ||
location * | l2, | ||
double * | localtheta, | ||
int | distance_metric | ||
) |
core_dcmg - Generate covariance matrix A in dense format between two sets of locations (l1, l2) (Matern Kernel). The routine makes only one pass through the tile A. One tile operation.
[out] | A | The m-by-n matrix on which to compute the covariance matrix. |
[in] | m | The number of rows in the tile A. |
[in] | n | The number of cols in the tile A. |
[in] | m0 | Global row index of the tile A. |
[in] | n0 | Global col index of the tile A. |
[in] | l1 | Location struct of the first input. |
[in] | l2 | Location struct of the second input. |
[in] | localtheta | Parameter vector that is used to generate the output covariance matrix. |
[in] | distance_metric | Distance metric "euclidean Distance (ED) ->0" or "Great Circle Distance (GCD) ->1" |
Definition at line 95 of file core_dcmg.c.
double core_dmdet | ( | double * | A, |
int | m, | ||
int | n, | ||
int | m0, | ||
int | n0 | ||
) |
core_dmdet - Calculate the determinant of the matrix A. The routine makes only one pass through the tile A. One tile operation.
[out] | A | The m-by-n matrix on which to calculate the determinant. |
[in] | m | The number of rows in the tile A. |
[in] | n | The number of cols in the tile A. |
[in] | m0 | global row index of the tile A. |
[in] | n0 | global col index of the tile A. |
Definition at line 51 of file core_dmdet.c.
void core_dzcpy | ( | double * | Z, |
int | m, | ||
int | m0, | ||
double * | r | ||
) |
core_dzcpy - Copy vector r to vector Z (one tile). The routine makes only one pass through the tile Z. One tile operation. One vector tile.
[out] | Z | The m-by-1 matrix on which to store r. |
[in] | m | The number of rows in the tile A. |
[in] | m0 | global row index of the tile A. |
[in] | r | double vector. |
Definition at line 49 of file core_dzcpy.c.
double core_ddotp | ( | double * | Z, |
double * | dotproduct, | ||
int | n | ||
) |
core_ddopt - Calculate the dot product of the Z vector. The routine makes only one pass through the tile A. One tile operation.
[in] | Z | The n-by-1 vector. |
[out] | dotproduct | dot product value. |
[in] | n | The number of rows in the tile A. |
Definition at line 43 of file core_ddotp.c.