Readout Functions

The data can be printed with std::cout:

template<class T> std::ostream &operator<<(std::ostream &stream, const danceq::internal::Vector<T> vector)

std::ofstream for the Vector class.

The function compatible with MPI. It has to be called by all processes.

Usage:

std::cout << vector: << vector << std::endl;

Parameters:

Returns:

std::ostream&

int32_t danceq::internal::Vector::print(void) const

Prints the vector.

T *danceq::internal::Vector::get_data_ptr(const uint64_t offset = 0UL)

Pointer to the local data with an offset.

Points to the local data with an offset*.

Note

If MPI is enabled, each rank points to its part with the same offset*.

const T *danceq::internal::Vector::get_const_data_ptr(const uint64_t offset = 0UL) const

const Pointer to the local data with an offset.

Points to the local data with an offset*. This allows only read access.

Note

If MPI is enabled, each rank points to its part with the same offset*.

int32_t danceq::internal::Vector::info(void) const

Prints information about the vector.

const T danceq::internal::Vector::operator[](const uint64_t index) const

Overloaded [] operator for readout.

Returns value at index* that can range from 0 to dim. The function is compatible with MPI and each processes obtains the same value stored at index*.

Note

It is not possible to set an entry with this operator! Use set_index(index,value) to do this.

Warning

The performance is bad if multiple values are requested, use get_const_data_ptr(…) instead.

std::vector<T> danceq::internal::Vector::gather_data(const int32_t root = 0) const

Gathers the data into one vector.

This is equivalent get_data() without MPI. The input root* is meaningless in this case.