Continuity

Definition

A function is continuous at a point when the surrounding input-output behavior agrees with the value at that point.

Simple form:

if x approaches a
then f(x) approaches f(a)

More compactly:

lim x->a f(x) = f(a)

In plain language:

the value at the point matches what nearby values approach

Continuity means the point fits the surrounding behavior of the function.

---

Simple Example

Function:

f(x) = x + 1

If x moves slightly, the output also moves slightly.

Example:

x = 2.00 -> f(x) = 3.00
x = 2.01 -> f(x) = 3.01
x = 1.99 -> f(x) = 2.99

The output does not suddenly jump somewhere else.

As x approaches 2, f(x) approaches 3.

Since:

f(2) = 3

the surrounding behavior matches the actual value at the point.

The function is continuous at x = 2.

---

What Does Nearby Mean?

Nearby does not mean a fixed distance like:

0.1
1
10

It means the input can be made close enough to control the output.

The idea is:

If you want the output to stay very close to f(a),
there is some small enough input range around a that guarantees it.

So continuity is not about one universal distance.

It is about controllable closeness.

Nearby means sufficiently close for the level of output closeness we care about.

---

Continuity Is Local

Continuity is checked at a specific point.

At x = a, the question is:

as inputs get closer to a,
do outputs get closer to f(a)?

It is not only about the previous input in a sequence.

It is about all inputs sufficiently close to that point.

Continuity is local agreement between a point and its surrounding behavior.

---

Continuity Does Not Mean Constant Change

Continuity does not mean the output changes at a constant rate.

Constant rate example:

f(x) = 2x + 1

This is a straight line.

A continuous function can also be curved:

f(x) = x^2

This function is continuous, but its rate of change is not constant.

Near x = 1, it changes more slowly.

Near x = 100, it changes much faster.

Continuous does not mean linear, flat, or constant rate.

---

Large Change Is Not Automatically A Jump

A function can be continuous and still change very quickly.

Example:

f(x) = 10000x

If input increases by 0.1, output increases by 1000.

That is a large change, but not a jump.

The function is still continuous because the change is predictable and traceable.

There is no break in the graph.

Large change means steepness; jump means broken behavior at a point.

---

Discontinuity

A discontinuity happens when the function breaks local input-output agreement.

Common forms:

jump
hole
undefined point
vertical break

Example:

f(x) = 0 if x < 0
f(x) = 1 if x >= 0

Near x = 0, the input can move by a tiny amount, but the output jumps from 0 to 1.

The surrounding behavior does not approach one shared value.

A discontinuity is a break in local behavior.

---

Discontinuity Does Not Mean Discrete

Discrete and discontinuous are different ideas.

discrete = possible values are separate
discontinuous = function behavior breaks over a space where closeness matters

Example:

f(x) = 0 if x < 0
f(x) = 1 if x >= 0

The input x can be continuous because it can be any real number.

But the function is discontinuous at x = 0.

Discrete describes the set of values; discontinuity describes broken local behavior.

---

Why Continuity Matters For Learning

If a function changes continuously, small parameter changes usually produce small output changes.

That makes improvement easier to reason about.

Example:

slightly change weight -> slightly change score

If outputs jumped unpredictably, small adjustments would be hard to control.

Continuity helps make gradual adjustment possible.

---

Final Summary

• Continuity at a means lim x->a f(x) = f(a).
• The value at the point must match what surrounding values approach.
• Nearby does not mean a fixed distance; it means sufficiently close.
• Continuous does not mean linear, flat, or constant rate.
• A large output change can be steepness, not necessarily a jump.
• Discontinuity is a jump, hole, break, or undefined behavior.
• Discontinuity does not mean discrete.
• Continuous quantities are easier to adjust gradually.
• This prepares the way for derivatives.