字典错误-朱庇特笔记本- Python 3

Question

这个功能以前运行得很好，现在我看不出它出了什么问题。下面是按这个顺序排列的函数、错误和hypJson字典。即使我注释掉这个特定的部分，在执行相同任务的其他部分也会出现相同的错误。任何帮助都将不胜感激。谢谢!

​

for idx, row in dfT1.iterrows():
    hypJson = json.loads(row['hyperparameters'])
    if hypJson['dropout'] not in d1:
        d1[hypJson['dropout']] = (row['test_accuracy'] * len(row['test_labels'], len(row['test_labels']))
        t1count[hypJson['dropout']] = 1
    else:
        d1[hypJson['dropout']] = (d1[hypJson['dropout']][0] + row['test_accuracy'] * len(row['test_labels']), #correct
                             d1[hypJson['dropout']][1] + len(row['test_labels'])) #total
        t1count[hypJson['dropout']] = t1count[hypJson['dropout']] + 1

​

​

  File "<ipython-input-19-1326a1a48cb8>", line 11
    t1count[hypJson['dropout']] = 1
          ^
SyntaxError: invalid syntax

​

​

{'dropout': 0, 'optimizer': 'sgd-001-0.9-nesterov', 'deep-dense-top': False, 'convnet-freeze-percent': 0}
{'dropout': 0, 'optimizer': 'sgd-001-0.9', 'deep-dense-top': False, 'convnet-freeze-percent': 0}
{'dropout': 0, 'optimizer': 'adam', 'deep-dense-top': False, 'convnet-freeze-percent': 0}
{'dropout': 0.1, 'optimizer': 'sgd-001-0.9-nesterov', 'deep-dense-top': False, 'convnet-freeze-percent': 0}
{'dropout': 0.1, 'optimizer': 'sgd-001-0.9', 'deep-dense-top': False, 'convnet-freeze-percent': 0}
{'dropout': 0.1, 'optimizer': 'adam', 'deep-dense-top': False, 'convnet-freeze-percent': 0}

​

Answer 1

上面的一行有不平衡的括号：

d1[hypJson['dropout']] = (row['test_accuracy'] * len(row['test_labels'], len(row['test_labels']))
                                                                    ^^^^ here

它应该是：

d1[hypJson['dropout']] = (row['test_accuracy'] * len(row['test_labels']), len(row['test_labels']))

Python通常会在语法错误后面的行上报告语法错误，因为缺少的结束括号本身并不是语法错误，所以它一直扫描代码，直到找到肯定是语法错误的东西，比如您的示例中的(...) t1count：

... = (row['test_accuracy'] * len(row['test_labels'], len(row['test_labels']))
t1count[hypJson['dropout']] = 1
^^^^^^^^ this is seen as (stuff)t1count, which is a syntax error