17. Notations and common conventions¶
In all this documentation, we use standard notations of linear algebra, data assimilation (as described in [Ide97]) and optimization.
For algebraic formulations, vectors are written horizontally or vertically
without making a difference. Matrices can be written either normally as a
serie of number series, either in text condensed notation, consisting in the
use of a space to separate the values, and a “
;” to separate the rows of
the matrix, in a continuous line.
Files can be indicated by an absolute or relative path. For some old or shared file systems, the full name, including the path, must not contain more than 256 characters. In the case of Python files, it is advisable not to use dots in the name apart for the extension, to prevent difficulties in using it that are then complicated to diagnose.
File type and naming conventions rely heavily on the extensions of the files themselves. Some of them are briefly specified here, without being exhaustive:
.py: data or commands text file of Python type
.comm: commands text file of ADAO EFICAS type
.xml: data text file of XML type or commands (for YACS, not exclusive)
.txt: data text file with space separator
.dat: data text file with space separator
.csv: data text file with comma or semicolon separator
.tsv: data text file with tab separator
.npy: data binary file of type Numpy mono-variable
.npz: data binary file of type Numpy multi-variables
.sdf: data binary file of type Scientific Data Format
For the use of some of these files, one can refer to the keyword “DataFile” as Pseudo-types of digital data description.