Why Do We Assume Pᵢ = rⁱ While Solving Difference Equations?

When learning linear difference equations for the first time, many students wonder:

Why do we suddenly assume:

$P_i = r^i$

Is this just guessing?

The answer is no. This assumption is made because exponential-type functions behave very nicely under shifting operations.

The Original Equation

Consider the recurrence relation:

$pP_{i+1} - P_i + qP_{i-1} = 0$

Notice that the equation contains shifted terms:

$P_{i+1}$
$P_i$
$P_{i-1}$

To solve such equations efficiently, we want a function whose shifted versions remain closely related to itself.

Testing the Choice Pᵢ = rⁱ

Suppose we assume:

$P_i = r^i$

Then:

$P_{i+1} = r^{i+1}$

Using exponent rules:

$r^{i+1} = r \cdot r^i$

Therefore:

$P_{i+1} = rP_i$

Similarly:

$P_{i-1} = r^{i-1}$

which becomes:

$P_{i-1} = \frac{1}{r}P_i$

This is extremely useful because shifting the index only multiplies the function by constants.

The overall structure remains unchanged.

Why This Helps

Substituting the assumed solution into the recurrence relation:

$pP_{i+1} - P_i + qP_{i-1} = 0$

gives:

$pr^{i+1} - r^i + qr^{i-1} = 0$

Factor out the common term:

$r^{i-1}(pr^2 - r + q) = 0$

Now the complicated recurrence relation has transformed into an algebraic equation:

$pr^2 - r + q = 0$

This equation is called the characteristic equation.

Solving an algebraic equation is much easier than directly solving the recurrence relation.

Why Not Choose Something Else?

Suppose we try:

$P_i = i^2$

Then:

$P_{i+1} = (i+1)^2$

which expands to:

$i^2 + 2i + 1$

Now additional terms appear, and the structure becomes messy.

The shifted function is no longer a simple multiple of the original function.

This makes solving the equation difficult.

The Special Property of Exponential Functions

Exponential functions have a remarkable property:

Shifting the input does not change their overall form.

For example:

$r^{i+1} = r \cdot r^i$

and

$r^{i-1} = \frac{1}{r}r^i$

The same function keeps reappearing, only scaled by constants.

This “self-reproducing” behavior is exactly what makes exponential functions ideal for solving linear recurrence relations.

Connection with Differential Equations

A very similar idea appears in differential equations.

For differential equations, we often assume:

$y = e^{rx}$

because differentiation gives:

$\frac{d}{dx}(e^{rx}) = re^{rx}$

Again, the function reproduces itself after the operation.

Similarly:

derivatives preserve exponential form in differential equations
shifts preserve exponential form in difference equations

That is why:

$P_i = r^i$

is the natural discrete counterpart of:

$y = e^{rx}$

Deep Mathematical Insight

Operations such as:

differentiation
shifting

have special functions called eigenfunctions that remain essentially unchanged under the operation.

For:

derivatives → exponential functions $e^{rx}$
shifts → geometric sequences $r^i$

This is why exponential-type trial solutions appear repeatedly throughout mathematics, physics, probability, engineering, and computer science.

Final Intuition

We assume:

$P_i = r^i$

because:

shifted versions remain proportional to the original function
substitution simplifies the recurrence relation
the recurrence transforms into an algebraic equation
exponential functions naturally “fit” linear systems involving shifts

This idea forms the foundation of solving linear recurrence relations and difference equations.

Additional menu

Share this:

Like this: