So far, we’ve seen how we can show that limits exist using a delta-epsilon proof, or by changing coordinates. In single variable calculus, we were often able to evaluate limits by direct substitution. For example, we could evaluate
We are able to do this because the function is continuous. Recall that a function is continuous at if For continuous functions, we can evaluate limits by simply plugging in the value.
Once we define continuity for multivariable functions, and determine which functions are continuous, we can use similar methods to evaluate multivariable limits.
Continuity
We define continuity similarly to how we did in single variable calculus.
Also similar to single variable calculus, virtually all of the common functions that we work with are continuous on their domains. That is, anywhere that they’re defined, they are continuous.
- polynomials
- root functions
- rational functions
- trigonometric functions and inverse trigonometric functions
- exponential functions and logarithmic functions
Furthermore, all of the ways that we’d like to combine continuous functions will result in another continuous function.
If and are continuous at , then is also continuous at .
If and are continuous at , then is also continuous at .
If is continuous at , and is continuous at , then is continuous at .
Note that even if a function is discontinuous at a point, it’s still possible that the limit exists. In this case, you’ll need to use a change of coordinates or other method to evaluate the limit.
Limits in General
So far, we’ve defined limits of scalar-valued functions, . We’ve seen how we can evaluate these limits, or show that they do not exist. However, we’ve yet to deal with more general multivariable functions, .
Fortunately, limits in the more general setting turn out to be an easy extension of the limits that we’ve already defined. That is, if we have a function , we can write in terms of its coordinate functions, Then, we can use the limits of the coordinate functions to define a limit of .
So, we can evaluate the limit of a function by taking the limit of the component functions. Because of this, the results and methods that we’ve used for limits of scalar functions carry naturally over to this more general setting.