Python中使用爱因斯坦求和约定进行张量收缩
对于使用爱因斯坦求和约定的张量收缩,请在Python中使用numpy.einsum()方法。第一个参数是下标,它指定用于求和的下标,以逗号分隔的下标标签列表表示。第二个参数是操作数,即操作的数组。
einsum()方法对操作数评估爱因斯坦求和约定。使用爱因斯坦求和约定,许多常见的多分量线性代数数组操作可以以简单的方式表示。在隐式模式下,einsum计算这些值。
在显式模式下,einsum通过禁用或强制对指定的下标标签进行求和,提供了进一步的灵活性来计算可能不被认为是经典爱因斯坦求和运算的其他数组运算。
步骤
首先,导入所需的库:
import numpy as np
使用array()方法创建两个NumPy一维数组:
arr1 = np.arange(60.).reshape(3,4,5) arr2 = np.arange(24.).reshape(4,3,2)
显示数组:
print("Array1...\n",arr1) print("\nArray2...\n",arr2)
检查两个数组的维度:
print("\nDimensions of Array1...\n",arr1.ndim) print("\nDimensions of Array2...\n",arr2.ndim)
检查两个数组的形状:
print("\nShape of Array1...\n",arr1.shape) print("\nShape of Array2...\n",arr2.shape)
对于使用爱因斯坦求和约定的张量收缩,请在Python中使用numpy.einsum()方法:
print("\nResult (Tensor contraction)...\n",np.einsum('ijk,jil->kl', arr1, arr2))
示例
import numpy as np # Creating two numpy One-Dimensional array using the array() method arr1 = np.arange(60.).reshape(3,4,5) arr2 = np.arange(24.).reshape(4,3,2) # Display the arrays print("Array1...\n",arr1) print("\nArray2...\n",arr2) # Check the Dimensions of both the arrays print("\nDimensions of Array1...\n",arr1.ndim) print("\nDimensions of Array2...\n",arr2.ndim) # Check the Shape of both the arrays print("\nShape of Array1...\n",arr1.shape) print("\nShape of Array2...\n",arr2.shape) # For Tensor contraction with Einstein summation convention, use the numpy.einsum() method in Python. print("\nResult (Tensor contraction)...\n",np.einsum('ijk,jil->kl', arr1, arr2))
输出
Array1... [[[ 0. 1. 2. 3. 4.] [ 5. 6. 7. 8. 9.] [10. 11. 12. 13. 14.] [15. 16. 17. 18. 19.]] [[20. 21. 22. 23. 24.] [25. 26. 27. 28. 29.] [30. 31. 32. 33. 34.] [35. 36. 37. 38. 39.]] [[40. 41. 42. 43. 44.] [45. 46. 47. 48. 49.] [50. 51. 52. 53. 54.] [55. 56. 57. 58. 59.]]] Array2... [[[ 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 Array1... 3 Dimensions of Array2... 3 Shape of Array1... (3, 4, 5) Shape of Array2... (4, 3, 2) Result (Tensor contraction)... [[4400. 4730.] [4532. 4874.] [4664. 5018.] [4796. 5162.] [4928. 5306.]]
广告