5
$\begingroup$

I am tring to solve equation $$\frac{2 \log (x y+2021)}{\log (2)}+4^{x+1}=2^{x y+2022}+\frac{2 \log (2 x+1)}{\log (2)},$$ where $x$, $y$ are integer numbers. I tried

Reduce[4^(x + 1) + 2 Log[2, x y + 2021] == 
  2^(x y + 2022) + 2 Log[2, 2 x + 1], {x, y}, Integers]

My computer runs too long. How can I reduce timing to solve this equation?

$\endgroup$
2
  • $\begingroup$ This is math, not Mathematica. I vote to close the question. $\endgroup$
    – user64494
    Commented Jul 14, 2020 at 17:23
  • $\begingroup$ @user64494 Yes, I solved it by hand. But, I want to solve it by Mathematica. $\endgroup$ Commented Jul 15, 2020 at 0:24

1 Answer 1

6
$\begingroup$

For me, neither Reduce or Solve seem to be able to provide any solutions. However, if we fix x to integer values, FindInstance is able to give us at least a solution. For the first couple of integers we get

{x -> 1, y -> -2018}
{x -> 2, y -> -1008}
{x -> 3, y -> -2014/3}
{x -> 4, y -> -503}
{x -> 5, y -> -402}
{x -> 6, y -> -1004/3}
{x -> 7, y -> -2006/7}

There is a pattern here, and one can check that the following is always solution to your equation

y[x_] := -2020/x + 2 

This can be checked analytically

{4^(x + 1) + 2 Log[2, x y + 2021],
  2^(x y + 2022) + 2 Log[2, 2 x + 1]} /. y -> y[x] // Simplify

(*
   {4^(1 + x) + (2 Log[1 + 2 x])/Log[2], 
    4^(1 + x) + (2 Log[1 + 2 x])/Log[2]}
*)

so this also holds for real values of x.

For y to be an integer, x needs to be a divisor of 2020, so we get the following set of integer solutions

{{x -> -2020, y -> 3}, {x -> -1010, y -> 4}, {x -> -505, 
  y -> 6}, {x -> -404, y -> 7}, {x -> -202, y -> 12}, {x -> -101, 
  y -> 22}, {x -> -20, y -> 103}, {x -> -10, y -> 204}, {x -> -5, 
  y -> 406}, {x -> -4, y -> 507}, {x -> -2, y -> 1012}, {x -> -1, 
  y -> 2022}, {x -> 1, y -> -2018}, {x -> 2, y -> -1008}, {x -> 4, 
  y -> -503}, {x -> 5, y -> -402}, {x -> 10, y -> -200}, {x -> 20, 
  y -> -99}, {x -> 101, y -> -18}, {x -> 202, y -> -8}, {x -> 404, 
  y -> -3}, {x -> 505, y -> -2}, {x -> 1010, y -> 0}, {x -> 2020, 
  y -> 1}}

However, there may be solutions outside of this set.

Edit

Your equation can be rewritten as \begin{equation} 2^{2x+1}-2^{xy+2021} = \log_2{\left(\frac{2x+1}{xy+2021}\right)} \end{equation}

The solution found above is just the condition for both sides to vanish, i.e.

\begin{equation} 2x+1=xy+2021, \end{equation}

which is a much nicer derivation.

$\endgroup$
4
  • 3
    $\begingroup$ It should be noticed that 4^(x + 1)- 2^(x y + 2022) is a rational number assuming x and y integer. This implies a certain condition for the arguments of the logarithms. BTW, all the solutions found with Maple satisfy both 4^(x + 1)== 2^(x y + 2022) and 2 Log[2, x y + 2021] == 2 Log[2, 2 x + 1]. $\endgroup$
    – user64494
    Commented Jul 14, 2020 at 14:02
  • $\begingroup$ Up to traditional math, {x -> -20, y -> 103} is not a solution in view of FunctionDomain[4^(x + 1) + 2 Log[2, x y + 2021], {x, y}, Reals] which produces x y > -2021. The true solutions are found by Reduce[4^(x + 1) == 2^(x y + 2022) && Log[2, x y + 2021] == Log[2, 2 x + 1], {x, y}, Integers]. $\endgroup$
    – user64494
    Commented Jul 15, 2020 at 4:06
  • $\begingroup$ I don't understand your point. At {x -> -20, y -> 103} both sides of the original equation become complex, but we have equality nonetheless, so it is a valid integer solution. $\endgroup$
    – Hausdorff
    Commented Jul 15, 2020 at 11:39
  • $\begingroup$ Hausdorff, up to traditional math, integer solutions have to belong to the domains of the functions which are defined by the LHS and RHS over the reals and Reduce[Log[2, x y + 2021] == Log[2, 2 x + 1], {x, y}, Integers] takes it into account. Don't hesitate to ask for further explanation in need. $\endgroup$
    – user64494
    Commented Jul 15, 2020 at 18:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.