Source

Adapted from my own handwritten and hand-typed notes from Year 1 to Year 3 of my EEE degree

We say when gets as close to as we like whenever is close enough to , along any possible path. This "any path" part is what makes 2D limits harder than 1D limits, where you only approach from the left or the right.

Formal definition

For every there is a such that

If the limit exists, it is unique.

Limit laws

If both and exist at :

  • , provided
  • for non-negative integers

Continuity

is continuous at if

Polynomials and rational functions (where the denominator is not zero) are continuous on their domains, so for them the limit is just direct substitution.

Method for evaluating a limit

  1. Substitute . A finite number is the answer. means the limit does not exist. A form like means keep going
  2. Factorise and cancel the factor causing
  3. Rationalise when a square root difference causes trouble
  4. Path test to show the limit does not exist
  5. Squeeze theorem to show the limit does exist

No L'Hopital

L'Hopital's rule does not apply to limits in two variables.

Path test

Substitute a path through (for try , , , , ) and take a one-variable limit.

Conclusions:

  • If one path gives (or no limit), the 2D limit does not exist
  • If two paths give different values, the 2D limit does not exist
  • If all paths give the same value, this proves nothing, so switch to the squeeze theorem

Squeeze theorem

If near and , then . When the expected limit is , bound from above, since is automatic.

Worked example: limit does not exist

Two paths:

  • Along :
  • Along :

Two paths give different values, so the limit does not exist. Since but there is no limit, is not continuous at the origin.

Worked example: limit exists

Find . Paths all give , so try squeezing. Since ,

so the limit is . The next idea built on limits is the partial derivative.