To demonstrate Lorentz transformations mathematically, we assume that time is a dimension (via linear transformations, etc.), what physical argument requires us to do this?
Details for the reason for this question: we suppose two observers O and O' at a point M, and a lamp which turns on and off every second for a short period of time which is very very small compared to the second and a distance L of the two observers, if O' begins to move away from the source at the moment it receives a signal, a simple calculation gives
$$ct'=L+vt'$$ $$ct'=L'=\frac{L}{1-\beta}\;\;,\;\;\beta=v/c$$
i.e. $$L=ct'-vt'=k_{1}(L-vt)\;\;,\;\; k_{1}=\frac{1}{1-\beta}$$ if the observer O ' approaches the source, we find $$L=ct''+vt''=k_{2}(L+vt)\;\;,\;\; k_{2}=\frac{1}{1+\beta}$$ if the observer O ' moves away orthogonally to the radius vector of the light wave, we find ((Pythagoras) $$L^{2}=\gamma^{2}L^{2}(1-\beta^{2})=\gamma(L-\beta L).\gamma (L+\beta L)$$ $$L^{2}=\gamma(L-vt).\gamma (L+vt)=L_{1}L_{2}$$
with $$\begin{cases} L_{1}=\gamma(L-vt) \\ L_{2}=\gamma (L+vt) \end{cases}$$ they have the same form as the Lorentz transformations but have nothing to do with the observer O' !