data_vector
In: Data Management
The data_vector is an entity with two faces. First, it is a memory representation of uniformly structured data (as e.g. set of several load regimes). Data used in PragTic are all of a native character – if they are taken from elsewhere, they are converted into at least one data_vector in the data_base. The notion “at least” is related to the fact, that not all data in FEM are structured to an uniform length – the description of an elements topology with various element types used is a good example here. Two data_vectors are build in such a case. First of them contains headers of elements (theirs IDs, material properties group etc.) and pointers to positions inside the other data_vector, where element’s nodes are specified.
Once converted into or build in the data_base, the data_vector has also a physical representation. This is an independent file utilising the same rootname as the task named by the user. The move of data from the hard disc to memory and back is cached. Only a part of data from the file is read to the memory. When another part of data is necessary, the data in cache is saved to the file and a new cache is read. Any potential loss of data is thus minimized.
To keep the order among data_vectors in the data_base a unique ID-name is assigned to each of them.
Each file / data_vector contains a header except of the raw data. Here an information about its content is preserved.
More:
© PragTic, 2007
This help file has been generated by the freeware version of HelpNDoc