Table 4.2.1 lists the four second-order partial derivatives for a function two variables. There are three common styles of notation for these derivatives, the subscript notation, the operator notation, and the D-operator notation.
Notation
|
Second-Order Partial Derivatives for
|
Subscript
|
|
Operator
|
|
D-operator
|
|
Table 4.2.1 Second-order partial derivatives for
|
|
|
For subscript notation, the lexical order of the subscripts is the order in which the derivatives are taken. Thus, in the derivative , the derivative with respect to is taken first, while the derivative with respect to is taken second.
However, for operator notation, the operators are applied from the left, so that in the derivative , the derivative with respect to is taken first, with the derivative with respect to taken second. Hence
and
The subscripts in D-operator notation refer to the variable in that position in the argument list for the function. This, for , the subscript 1 refers to the first variable, namely, ; whereas the subscript 2 refers to the second, namely, . Thus
and
At any point at which is sufficiently well behaved, the mixed partial derivatives and are equal. Section 4.11 explores the precise conditions under which the mixed partial derivatives are equal. Example 4.2.5 explores a function for which the mixed partials are not equal at the origin.
Table 4.2.2 lists the eight third-order partial derivatives for .
|
|
|
|
|
|
|
|
Table 4.2.2 Third-order partial derivatives for
|
|
|
Under the assumption of the equality of mixed partials, the following equalities hold.
|
|
|
|
|
|
|
|
The Calculus palette contains templates for both first-order and second-order partial derivatives of a function of two variables. These templates can be edited to fit other cases, and the symbol , found in the Operators palette, can be used to build operator-notation templates.
Maple's D-operator acts on a function and returns derivatives as functions. Hence, parentheses are required, as in and . In the first instance, the mixed partial is returned as a function; in the second, it is evaluated at the point , so is simply an expression in and .