x_{1} is directly/linearly proportional to x_{2} via coefficient/constant of proportionality x_{3} and background/constant offset x_{4}; there exists constants x_{3} and x_{4} such that x_{1 }= x_{3×x}.

x_{2} may be an incomplete list of quantities. These should be variable. x_{3} should be a fixed constant for the universe, situation, object in question. x_{4} is possibly variable but may also be fixed by the environmwnt/context/situation. The equality should be exact in a model/in theory. If one is taking a linear approximation of a function which does not behave linearly with respect to x_{2}, then caveats must be included (such as by using a tanru in order to specify approximation) or by separation (using be; "f approximates g, which is proportional [tartidu] to h, with error z, which is proportional [tartidu] to w"). x_{2} can be any of a fairly general class of nunvers, including a variable/unknown (with or without units) and a function (which need not be linear). Inverse or joint proportionality can simply be expressed using division and multiplication respectively. All terbri should be filled with either pure numbers (li; general) or quantified concepts (such as physical quantities).