In order for source code to be readable and maintainable, variable names should be expressive, i.e. they should imply what the variable represents. By doing that, documenting variable names becomes virtually unnecessary.
However, in scientific programming we often need to deal with fairly complex mathematical equations. Then it is tempting to use the same or at least similar symbols as we find in the equation in the text book. While these symbols are obvious to people with domain knowledge, for programmers focusing on software optimization and industrialization these symbols are often hard to read.
Out of these considerations we decided that in pyLife for physical quantities the variable names as described in this document are mandatory. For physical quantities not described in this document, you can either use an expressive variable name or you can document a symbol in your module documentation.
Roman letters can be used as is, capital or small. Greek letters could actually
also be written as unicode letters. Yes, x = x_0 * np.exp(-δ*t) * np.cos(ω*t +
φ) is perfectly valid Python code. However, not all of us are using decent
systems which allow you to type them easily. That's why for Greek letters we
would spell them out like alpha. This does not work for lambda, though
as it is a keyword in python.
Indices should be separated with an underscore (_) from the symbol.
However, in some cases the underscore is not uses (see below.)
- Stress tensor variables:
S11,S22,S33,S12,S13,S23 - Cyclic stress variables:
*
amplitude: stress or load amplitude, *meanstress: meanstress, *R: R-value
- Cartesian coordinates:
x,y,z
- Displacements in a Cartesian coordinate system:
dx,dy,dz
SDendurance limit in load direction,SD_xxforxxpercent failure probabilityNDendurance limit in cycle direction,ND_xxforxxpercent failure probabilityTSscatter in load direction (=SD_10/SD_90)TNscatter in load direction (=ND_10/ND_90)kslope of the Wöhler curve,k_1above the endurance limit,k_2below the endurance limit