Question

我是Python的新手（使用3.5和Anaconda） - 以前在MATLAB方面的经验。非常感谢您的帮助。如果有更简单的方法，请告诉我。

我从一些实验设备的pdf文件中读取并清理了数据并将其附加到一个列表中：

>print(outputdata)

[[['2.37701'], ['-'], ['-'], ['-'], ['-'], ['18.95276'], ['5.07365e-1']], [['2.75613'], ['-'], ['-'], ['-'], ['-'], ['16.99642'], ['4.10023e-1']], [['1.80527'], ['-'], ['-'], ['-'], ['-'], ['20.75384'], ['4.58238e-1']], [['1.58721'], ['-'], ['-'], ['-'], ['-'], ['18.06942'], ['3.81128e-1']], [['1.98336'], ['-'], ['-'], ['-'], ['-'], ['18.20776'], ['3.64733e-1']], [['1.75710'], ['-'], ['-'], ['-'], ['-'], ['23.03760'], ['4.36234e-1']], [['1.58967'], ['-'], ['-'], ['-'], ['-'], ['21.43884'], ['3.88509e-1']], [['2.37701'], ['-'], ['-'], ['-'], ['-'], ['18.95276'], ['5.07365e-1']]]

我正在尝试从列表的每个元素中获取每个元素，并将其保存到新列表中。我还想清理数据以删除括号和引号并保留数字。我需要对此进行操作，所以我计划转换为numpy数组，然后将其添加到DataFrame以更轻松地导出到Excel（我已经有了导出代码）。每个列向量对应于特定标题：

Molecule = ["H2", "Ar", "Methane", "Ethane", "Ethylene", "Propane(C3H8)", "Propylene"]

以下是所需H2数据的示例：

我首先做到了这一点：

outputdatalist = [x[0] for x in outputdata]

具有以下输出：

[['2.37701'], ['2.75613'], ['1.80527'], ['1.58721'], ['1.98336'], ['1.75710'], ['1.58967'], ['2.37701']]

然后

for row in outputdatalist:
    print(' '.join(row))  # I need to append this on every iteration

我的untythonic（并且不成功）这样做的方法是按如下方式执行double（triple？）for循环：

outputdatalist = []
for counter, elem in enumerate(Molecule):
for counter1, elem1 in enumerate(outputdata):
     outputdatalist[counter] = [x[counter1] for x in outputdata]

然后将每个outputdatalist [i]转换为np数组，然后使用类似

的循环遍历pd.Dataframe

pd.DataFrame({Molecule[i]: outputdatalist[i]})

Answer 1

您可以使用nested list comprehension，这似乎更快作为apply的解决方案：

df = pd.DataFrame([[y[0] for y in x] for x in outputdata], columns=Molecule)
print (df)
        H2 Ar Methane Ethane Ethylene Propane(C3H8)   Propylene
0  2.37701  -       -      -        -      18.95276  5.07365e-1
1  2.75613  -       -      -        -      16.99642  4.10023e-1
2  1.80527  -       -      -        -      20.75384  4.58238e-1
3  1.58721  -       -      -        -      18.06942  3.81128e-1
4  1.98336  -       -      -        -      18.20776  3.64733e-1
5  1.75710  -       -      -        -      23.03760  4.36234e-1
6  1.58967  -       -      -        -      21.43884  3.88509e-1
7  2.37701  -       -      -        -      18.95276  5.07365e-1

计时 :(小型数据框）

In [21]: %timeit pd.DataFrame([[y[0] for y in x] for x in outputdata], columns=Molecule)
1000 loops, best of 3: 1.04 ms per loop

In [22]: %timeit (pd.DataFrame(outputdata, columns=Molecule).apply(lambda x: x.str[0]))
100 loops, best of 3: 4.59 ms per loop

Pythonic方式的多个for循环，在每次迭代时创建新列表，还清理数据？

1 个答案: