I'd like to ask how to get the parameterization to this equation: $3z_1^2+z_2^2=156$, where $z_1$ and $z_2$ are both Gaussian integers.
More generally, is there any parameterization to the general equation $az_1^2+bz_2^2=c$?
I'd like to ask how to get the parameterization to this equation: $3z_1^2+z_2^2=156$, where $z_1$ and $z_2$ are both Gaussian integers.
More generally, is there any parameterization to the general equation $az_1^2+bz_2^2=c$?
Here is a discussion around solving $$3x^2+y^2=156\ ,\qquad x,y\in\Bbb Z[i]\ ,$$ based on the factorization of $156$ in the ring $R$ of integers of the number field $$K=\Bbb Q(\sqrt{-1},\sqrt{-3})=\Bbb Q(i,\sqrt 3)\ .$$ I am using $x,y$ instead of the indexed $z$-variables, since it is comfortable to use further the implicit notations $x=x_1+ix_2$, $y=y_1+iy_2$, and avoiding too much indices makes typing simpler.
Note that any solution of the given equation can be rewritten $$ (y+x\sqrt{-3}) (y-x\sqrt{-3}) =156\ .$$ So we have to find among all factors of $156\in R=\mathcal O_K$ the suitable ones (that have coefficients in $\Bbb Z$ w.r.t. the basis $\{1,i,\sqrt 3,i\sqrt{3}\}$). Note that $R=\mathcal O_K$ is a unique factorization domain. This, and some further steps can be get using sage, the sage code is generally object oriented, structural, and easily readable for mathematicians, i hope its application here by my typing hand does not suggest the contrary, the fields $Q=\Bbb Q(i)$, $L=Q(\sqrt{-3})$, and $K=\Bbb Q(i+\sqrt3)\cong L$ will be introduced, and some properties of them are questioned...
Let us also ask for the factors of $156\in \mathcal O_L$, and their multiplicities...
Some results were manually rearranged to fit in the shown width.
Humanly, this means we have the following factors that appear in the decomposition of $156$:
$$ \begin{aligned} f_1 &=\frac 12(1+i+\sqrt{-3}+i\sqrt{-3}) &&\sim \bar f_1\ , &&\text{ order }4\ ,\\ f_2 &=\sqrt{-3}&&\sim \bar f_2\ , &&\text{ order }2\ ,\\[2mm] f_3 &=\frac 12(3-\sqrt{-3})-i &&\sim \bar f_5\ , &&\text{ order }1\ ,\\ f_4 &=\frac 12(3-\sqrt{-3})+i&&\sim \bar f_6\ , &&\text{ order }1\ ,\\[2mm] f_5 &=\frac 12(-3+\sqrt{-3}(1-i)+i)&&\sim \bar f_3\ , &&\text{ order }1\ ,\\ f_6 &=\frac 12(-3+\sqrt{-3}(1+i)-i)&&\sim \bar f_4\ , &&\text{ order }1\ , \end{aligned} $$ and now we have to group them w.r.t. the Galois conjugation which maps $i\to i$, $\sqrt{-3}\to-\sqrt{-3}$. A solution of $y^2+3x^2=156$ in integers corresponds to a solution for the relative norm problem $$ N_{L:\Bbb Q(i)}(y+x\sqrt{-3})=(y+x\sqrt{-3})(y-x\sqrt{-3})=156\ , $$ and the two factors contain $(L:\Bbb Q(i))$-conjugated prime factors. These factors are those from above, there may also appear a unit, and we have a classification for the unit group of the field. It has a $12$--torsion part generated by the fourth roots of the unit $1,i,-1,-i$, and the sixth roots of the unit, $1,u,u^2,u^3=-1, -u,-u^2$, where $u=\frac 12(1+\sqrt{-3})$, and the two groups of order $4$, resp. $6$ have the obvious subgroup of order two, $\pm 1$ in common. It also has a free part, in the code we have the generating unit $$ W = 1+\frac i2(1-\sqrt{-3})\ , $$ and its first few powers are:
Not all of them have coefficients in $\Bbb Z[i]$, but $W^{6k}\in\Bbb Z[i]$ is a fact.
Now we can proceed as follows. We find all the solutions using a unit and some (suitably chosen) factors among $f_1,\dots, f_6$. These provide a finite set of solutions. We multiply them with powers of $W$ and get algebraic integers $y+x\sqrt{-3}$ of norm $156$. So the general solution is of the form: $$ y + x\sqrt{-3} = u\cdot f_1^2\, f_2\, f_{3\text{ or }5}\, f_{4\text{ or }6}\cdot W^k\ , $$ where $u$ is a unit (among $12$) and $k\in \Bbb Z$. We can pass from $(x,y)$ to a solution with real part of $x,y$ both $>0$ (by using powers of $i$).
Here are some values for $x,y\in\Bbb Z[i]$, computed via sage:
Results:
The general case depends on the way to capture a bigger class number, the computation of units is relatively simpler to get calculated. But note that the units can accidentally a "very big height" and hard to compute in special cases.