Python NumPy에서 polynomial.legendre.legval3d() 메서드를 사용하면 점 (x, y, z)에서 3차원 르장드르(Legendre) 급수를 손쉽게 평가할 수 있습니다. 이 메서드는 x, y, z에서 대응하는 값들의 조합으로 이루어진 점들에 대해 다차원 다항식의 값을 계산하여 반환합니다.
만약 계수 배열 c가 3차원보다 낮다면, 형태(shape)에 1이 암묵적으로 추가되어 3차원으로 만들어집니다. 결과의 형태는 c.shape[3:] + x.shape가 됩니다.
legval3d() 메서드의 매개변수
첫 번째 매개변수: x, y, z
3차원 급수가 평가될 점 (x, y, z)를 나타냅니다. 이때 x, y, z는 반드시 동일한 형태(shape)를 가져야 합니다. 만약 x, y, z 중 하나라도 리스트(list)나 튜플(tuple)이라면 먼저 ndarray로 변환되며, ndarray가 아닌 경우에는 스칼라(scalar)로 취급됩니다.
두 번째 매개변수: c
계수 배열입니다. 다중 차수(multidegree) i, j, k인 항의 계수가 c[i, j, k]에 저장되도록 정렬되어 있습니다. 만약 c의 차원이 3보다 크다면, 남은 인덱스들은 여러 개의 계수 집합을 나타내는 데 사용됩니다.
단계별 구현 방법
먼저 필요한 라이브러리를 임포트합니다.
import numpy as np
from numpy.polynomial import legendre as L
3차원 계수 배열을 생성합니다.
c = np.arange(24).reshape(2,2,6)
배열을 화면에 출력합니다.
print("Our Array...\n",c)배열의 차원을 확인합니다.
print("\nDimensions of our Array...\n",c.ndim)배열의 데이터 타입을 확인합니다.
print("\nDatatype of our Array object...\n",c.dtype)배열의 형태(shape)를 확인합니다.
print("\nShape of our Array object...\n",c.shape)이제 legval3d() 메서드를 사용해 점 (x, y, z)에서 3D 르장드르 급수를 평가합니다.
print("\nResult...\n",L.legval3d([1,2],[1,2],[1,2],c))전체 예제 코드
import numpy as np
from numpy.polynomial import legendre as L
# 3차원 계수 배열 생성
c = np.arange(24).reshape(2,2,6)
# 배열 출력
print("Our Array...\n",c)
# 차원 확인
print("\nDimensions of our Array...\n",c.ndim)
# 데이터 타입 확인
print("\nDatatype of our Array object...\n",c.dtype)
# 형태(shape) 확인
print("\nShape of our Array object...\n",c.shape)
# legval3d() 메서드로 점 (x, y, z)에서 3D 르장드르 급수 평가
print("\nResult...\n",L.legval3d([1,2],[1,2],[1,2],c))
실행 결과
Our Array...
[[[ 0 1 2 3 4 5]
[ 6 7 8 9 10 11]]
[[12 13 14 15 16 17]
[18 19 20 21 22 23]]]
Dimensions of our Array...
3
Datatype of our Array object...
int64
Shape of our Array object...
(2, 2, 6)
Result...
[ 276. 39723.75]
위 실행 결과에서 볼 수 있듯이, 생성된 계수 배열은 int64 타입의 3차원 배열(형태: (2, 2, 6))이며, legval3d() 메서드를 통해 점 (1, 1, 1)과 (2, 2, 2)에서 각각 276.0과 39723.75라는 평가 결과를 얻을 수 있습니다.