偏導數者,乃多元函數 f ( x 1 , x 2 , . . . , x i , . . . , x n ) {\displaystyle f(x_{1},x_{2},...,x_{i},...,x_{n})} 於某變量 x i {\displaystyle x_{i}} 之變率。
計偏導數時,若微分於某變量 x i {\displaystyle x_{i}} ( ∂ f ∂ x i {\displaystyle {\frac {\partial f}{\partial x_{i}}}} ),則視其余變量 x j {\displaystyle x_{j}} ( j ∈ Z ∪ [ 1 , n ] i̸ {\displaystyle j\in \mathbb {Z} \cup [1,n]\not i} )者為常數,僅以 x i {\displaystyle x_{i}} 作變量運之。
設 U {\displaystyle U} 為一開區間且函數 f : U → R {\displaystyle f:U\to R} , f {\displaystyle f} 之偏微分於變量 x i {\displaystyle x_{i}} ,於點 a : ( a 1 , a 2 , . . . a i , . . . , a n ) ∈ U {\displaystyle a:(a_{1},a_{2},...a_{i},...,a_{n})\in U} 之偏導數定義為: ∂ ∂ x i f ( a 1 , a 2 , . . . , a i , . . . , a n ) = lim Δ x i → 0 f ( a 1 , a 2 , . . . , a i + Δ x i , . . . , a n ) − f ( a 1 , a 2 , . . . , a i , . . . , a n ) Δ x i {\displaystyle {\frac {\partial }{\partial x_{i}}}f(a_{1},a_{2},...,a_{i},...,a_{n})=\lim _{\Delta x_{i}\to 0}{\frac {f(a_{1},a_{2},...,a_{i}+\Delta x_{i},...,a_{n})-f(a_{1},a_{2},...,a_{i},...,a_{n})}{\Delta x_{i}}}}
使 z = x 2 + x y + y 2 {\displaystyle z=x^{2}+xy+y^{2}} 。
視 y {\displaystyle y} 作常數, ∂ z ∂ x = 2 x + y {\displaystyle {\frac {\partial z}{\partial x}}=2x+y} 。
視 x {\displaystyle x} 作常數, ∂ z ∂ y = 2 y + x {\displaystyle {\frac {\partial z}{\partial y}}=2y+x} 。
https://www.mathsisfun.com/calculus/derivatives-partial.html