Main content
Multivariable calculus
Course: Multivariable calculus > Unit 2
Lesson 3: Partial derivative and gradient (articles)Introduction to partial derivatives
What is the partial derivative, how do you compute it, and what does it mean?
What we're building to
- For a multivariable function, like
, computing partial derivatives looks something like this:
- This swirly-d symbol,
, often called "del", is used to distinguish partial derivatives from ordinary single-variable derivatives. Or, should I say ... to differentiate them. - The reason for a new type of derivative is that when the input of a function is made up of multiple variables, we want to see how the function changes as we let just one of those variables change while holding all the others constant.
- With respect to three-dimensional graphs, you can picture the partial derivative
by slicing the graph of with a plane representing a constant -value and measuring the slope of the resulting curve along the cut.
What is a partial derivative?
We'll assume you are familiar with the ordinary derivative from single variable calculus. I actually quite like this notation for the derivative, because you can interpret it as follows:
- Interpret
as "a very tiny change in ". - Interpret
as "a very tiny change in the output of ", where it is understood that this tiny change is whatever results from the tiny change to the input.
In fact, I think this intuitive feel for the symbol is one of the most useful takeaways from single-variable calculus, and when you really start feeling it in your bones, most of the concepts around derivatives start to click.
For example, when you apply it to the graph of , you can interpret this "ratio" as the rise-over-run slope of the graph of , which depends on the point where you started.
How does this work for multivariable functions?
Consider some function with a two-dimensional input and a one-dimensional output.
There's nothing stopping us from writing the same expression, , and interpreting it the same way:
can still represent a tiny change in the variable , which is now just one component of our input. can still represent the resulting change to the output of the function .
However, this ignores the fact that there is another input variable . The input space now has multiple dimensions, so we can change the input in many directions other than the -direction. For example, what about changing slightly by some small value ? Now if we re-interpret to represent the tiny change to the function that this shift brings about, we would have a different derivative .
Neither one of these derivatives tells the full story of how our function changes when its input changes slightly, so we call them partial derivatives. To emphasize the difference, we no longer use the letter to indicate tiny changes, but instead introduce a newfangled symbol to do the trick, writing each partial derivative as , , etc.
You read the symbol out loud by saying "the partial derivative of with respect to ".
Example: Computing a partial derivative
Consider this function:
Suppose I asked you to evaluate , the partial derivative with respect to , at the input .
"What? But I haven't learned how yet!"
Don't worry, it's mostly just the same mechanics as an ordinary derivative.
From the introduction above, you should know that this is asking about the rate at which the output of changes as we nudge the -component of the input slightly, perhaps moving from to .
Since we only care about movement in the -direction, we might as well treat the -value as a constant. In fact, we can just plug in ahead of time before computing any derivatives:
Now, asking how changes in response to a small shift in is just an ordinary, single-variable derivative.
Without pre-evaluating
Now suppose I asked you to find , but I didn't ask you to evaluate it at a specific point. In other words, you should give me new multivariable function which takes any point as its input and tells me what the rate of change of near that point is as we move purely in the -direction.
You can start the same way, treating the value as a constant. However, this time, you cannot plug in an actual constant value, like . Instead, pretend that is constant and take the derivative:
Or rather, since to emphasize that this is a multivariable function, we use the symbol instead of :
As a sanity check, you can plug in to see that we get the same result as above.
"So, what's the difference between and ? They seem to be used the same way."
Honestly, as far as I'm concerned, there's not really a difference between these operations. You could be pedantic and say one is only defined for single variable functions. But as far as intuition and computation go, they are one and the same, and the difference is just meant to clarify what type of function is being differentiated.
Interpreting partial derivatives with graphs
Consider this function:
Here is a video showing its graph rotating, just to get a feel for the three-dimensional nature of it.
Think about the partial derivative of with respect to , perhaps evaluated at the point .
In terms of the graph, what does the value of this expression tell us about the behavior of the function at the point ?
Treat as constant slice graph with plane
The first step when computing this value is to treat as a constant. Specifically, if we are limiting our view to what happens at the point , we should only look at the set of points where . In three-dimensional space, this set is plane perpendicular to the -axis, passing through the origin.
This plane , shown in white, slices into the graph of along a parabolic curve, shown faintly in red. We can interpret as giving the slope of a tangent line to this curve. Why? Because is a slight nudge in the -direction, the run, and is the resulting change in the -direction, the rise.
What about at that same point ? The points where also make up a plane, but this time it's a plane perpendicular to the -axis intersecting the point . This slices the graph along a new curve, and will give the slope of that new curve.
Phrasing and notation
Here are some of the phrases you might hear in reference to this operation:
- "The partial derivative of
with respect to " - "Del f, del x"
- "Partial f, partial x"
- "The partial derivative (of
) in the -direction"
Alternate notation
In the same way that people sometimes prefer to write instead of , we have the following notation:
A note about "del"
While it's common to refer to the partial symbol as "del", this can be confusing because "del" is also the name of the Nabla symbol , which we will introduce in the next article.
A more formal definition
Although thinking of or as really tiny changes in the value of is a useful intuition, it is healthy to occasionally step back and remember that defining things precisely requires introducing limits. After all, what specific small value would be? One one hundredth? One one millionth? ?
The point of calculus is that we don't use any one tiny number, but instead consider all possible values and analyze what tends to happen as they approach a limiting value. The single variable derivative, for example, is defined like this:
represents the "tiny value" that we intuitively think of as .- The
under the limit indicates that we care about very small values of , those approaching . is the change in the output that results from adding to the input, which is what we think of as .
Formally defining the partial derivative looks almost identical. If is a function with multiple inputs, here's how that looks:
Similarly, here's how the partial derivative with respect to looks:
The point is that , which represents a tiny tweak to the input, is added to different input variables depending on which partial derivative we are taking.
People will often refer to this as the limit definition of a partial derivative.
Reflection question: How can we think about this limit definition in the context of the graphical interpretation above? What is ? What does it look like for ?
Summary
- For a multivariable function, like
, computing partial derivatives looks something like this:
- This swirly-d symbol
, often called "del", is used to distinguish partial derivatives from ordinary single-variable derivatives. - The reason for a new type of derivative is that when the input of a function is made up of multiple variables, we want to see how the function changes as we let just one of those variables change while holding all the others constant.
- With respect to three-dimensional graphs, you can picture the partial derivative
by slicing the graph of with a plane representing a constant -value, and measuring the slope of the resulting cut.
Want to join the conversation?
- Please, please can you label the axes three dimensional graphs completely and
more clearly? Your site is the best! I am 75 and trying to revise my university
math of some 50 years ago!
Many thanks(106 votes)- The up axis is always z, and you can use the right hand rule to see which the other two are; your index finger will be pointing at the x-axis, and you're middle finger will be pointing at the y-axis. (Note: different software use different conventions for labeling the axes. Sometimes z represents depth and y represents height. But mac grapher - the software Grant is using here - always uses the convention I described above.)(3 votes)
- In Example 2, isn't cos(0*pi) = cos(0) = 1? And therefore the answer is non-negative?(22 votes)
- Where it says: the set of all points where y = 0 includes all points of the form (x,0,z). I believe this set is mistakenly called the xy-plane. It should be called the xz-plane.(8 votes)
- Hi,
I find these articles very useful, and I'd like to keep them at hand for future references. Are they available for download, in pdf or some other print-friendly format, somewhere? Thanks.(4 votes) - Why there is no a mission in multivariable funtions?(1 vote)
- It's more about learning the techniques, notation, and order of operations than plugging and cranking a bunch of examples. It is also an area of math that isn't really intended to learn purely online. This section is more of a nudge to help learn or better understand something you are currently learning in a 5th semester math course. Usually proceeded by MA 161,162 and MA261,262 or something similar. I'm just using Purdues course #s.(4 votes)
- I still see the gradient vary on the video graphical representation. Does not make sense.(1 vote)
- The explanation with the graph is hard to follow. I see no relation between the graph being shown and the equations being listed. I don't understand any of it.(1 vote)
- If you think back to Calculus 1 (or single-variable calculus), recall the the derivative of a function is equal to its slope at any point. If you don't understand that concept, it might be good to look back and review the section on derivatives. In this case, when we take a slice of the graph, the two-dimensional intersection of the graph and the plane looks like a single-variable function. Which variable the function is of depends on the orientation of the graph. If the graph is parallel to the x-axis, it looks like a function of x, and if the graph is parallel to the y-axis, the intersection looks like a function of y. The partial derivative is a way to find the slope in either the x or y direction, at the point indicated. By treating the other variable like a constant, the situation seems to simplify to something we can understand in terms of single-variable derivatives, which we learned in Calc 1. If you still do not understand, let me know, and we can try to work it out. Make sure you have learned previous lessons well.(3 votes)
- when was this published?(2 votes)
- They kicked off with the playlist on multivariable calculus 4 or 5 years ago. Not sure about this current article though.(1 vote)
- Just a nitpicky question about notation, is the "∂" symbol also called del like this article says? I thought that "∇" was del, and that the partial symbol was just a stylized "d"(1 vote)
- Per the Greek alphabet, ∂ is lowercase delta whereas ∇ is uppercase delta.(2 votes)
- 2 quick questions:
1) example 2, sin( 0 + pi) = 0... however I ran my calculator and got the answer .0548. I checked and my settings are in degrees, not radians.
2) In both my homework (Dynamic Atmosphere) and your examples, when deriving 'y' for example, why do we make '2x' or 'x^2' equal to zero. Why do we not bring them as a constant?(1 vote)- If they are stand alone such as x^2+2x+2xy^3. If you are taking the partial derivative with respect to y, you treat the others as a constant. The derivative of a constant is 0, so it becomes
0+0+2x(3y^2). You'll notice since the last one is multiplied by Y, you treat it as a constant multiplied by the derivative of the function. I.E d/dx 2x^2 is just 2(d/dx(x^2)) or 2(2x) or 4x. Hope this helps.(2 votes)