Cyclops Tensor Framework
parallel arithmetic on multidimensional arrays
|
Go to the source code of this file.
Namespaces | |
CTF_int | |
Functions | |
template<typename dtype > | |
const char * | CTF_int::get_fmt () |
return format string for templated type More... | |
template<> | |
const char * | CTF_int::get_fmt< float > () |
template<> | |
const char * | CTF_int::get_fmt< double > () |
template<> | |
const char * | CTF_int::get_fmt< int > () |
template<> | |
const char * | CTF_int::get_fmt< int64_t > () |
template<typename dtype > | |
void | CTF_int::parse_sparse_tensor_data (char **lvals, int order, dtype const *pmulid, int *lens, int64_t nvals, CTF::Pair< dtype > *pairs, bool with_vals) |
parse string containing sparse tensor into data More... | |
template<typename dtype > | |
char * | CTF_int::serialize_sparse_tensor_data (int order, int *lens, int64_t nvals, CTF::Pair< dtype > *pairs, bool with_vals, int64_t &str_len) |
serialize sparse tensor data to create string More... | |
template<typename dtype > | |
int64_t | CTF_int::read_data_mpiio (CTF::World const *dw, char const *fpath, char ***datastr) |
read sparse tensor data from file using MPI-I/O, creating string with one entry per line (different entries on each process) More... | |
template<typename dtype > | |
void | CTF_int::write_data_mpiio (CTF::World const *dw, char const *fpath, char *datastr, int64_t str_len) |
write sparse tensor data to file using MPI-I/O, from string with one entry per line (different entries on each process) More... | |