python - python 中进行敏感性分析的 SAlib 库出错

Question

我正在尝试使用 Sobol 方法进行敏感性分析。我总是遇到无法解决的错误。代码和结果如下。输入变量范围在问题中确定，模型的输出保存在 sens_out.txt 中

from SALib.sample import saltelli
from SALib.analyze import sobol
from SALib.test_functions import Ishigami
import numpy as np
import subprocess

problem = {
    'num_vars': 30,
    'names': ['x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7', 'x8', 'x9', 'x10',
              'x11', 'x12', 'x13', 'x14', 'x15', 'x16', 'x17', 'x18', 'x19',
              'x20', 'x21', 'x22', 'x23', 'x24', 'x25', 'x26', 'x27', 'x28',
              'x29', 'x30'],
    'bounds': [[2, 6],
               [20, 30],
               [600, 900],
               [800, 1300],
               [0.01, 0.05],
               [0.1, 0.2],
               [30, 35],
               [0.0015, 0.003],
               [0.001, 0.0015],
               [0.50, 0.70],
               [0.50, 0.70],
               [0.50, 0.70],
               [0.25, 0.50],
               [0.0, 0.25],
               [0.4, 0.5],
               [0.5, 0.7],
               [0.5, 0.7],
               [0.5, 0.7],
               [0.55, 0.7],
               [0.6, 0.7],
               [0.013, 0.017],
               [0.027, 0.032],
               [0.17, 0.21],
               [0.007, 0.012],
               [0.01, 0.016],
               [0.02, 0.04],
               [0.8, 1.2],
               [2.0, 2.5],
               [8, 12],
               [90, 110]]
}

param_values = saltelli.sample(problem, 1000, calc_second_order=True)

#np.savetxt("/home/omar/Desktop/param_values.txt", param_values)

Y = np.loadtxt("/home/omar/Desktop/sens_out.txt", float)

Si = sobol.analyze(problem, Y, print_to_console=True)

<小时/>

结果 = 错误消息

    Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.7/dist-packages/spyderlib/widgets/externalshell/sitecustomize.py", line 540, in runfile
    execfile(filename, namespace)
  File "/home/omar/Desktop/untitled0.py", line 58, in <module>
    Si = sobol.analyze(problem, Y, print_to_console=True)
  File "/home/omar/SALib/SALib/analyze/sobol.py", line 81, in analyze
    A,B,AB,BA = separate_output_values(Y, D, N, calc_second_order)
  File "/home/omar/SALib/SALib/analyze/sobol.py", line 164, in separate_output_values
    AB[:, j] = Y[(j + 1):Y.size:step]
ValueError: could not broadcast input array from shape (1000,14) into shape (14000)

Answer 1

Y 的尺寸是多少？要使用 sobol.analyze 函数，Y 必须是列向量。在这种情况下，您似乎传递的是矩阵而不是向量，这似乎是问题所在。

此外，如果您计算二阶指数，则样本大小将为 N(k+2)，或者在本例中为 32,000。如果模型仅返回 14,000 个值，这将导致数组维度的另一个问题。