csv文件在以浮点数编写时提供字符串

Question

我的csv文件如下所示：

"(-0.16192019, -0.007456851, 0.08841695)","(-0.10439591, 0.024501082, 0.12783173)","(-0.05965481, 0.049002163, 0.12144014)",...

但我不要引号

我写的功能为：

for j in range(len(sensor_data)):
    with open(names[j] + "P" + str(i) + ".csv", "wb") as f:
        writer = csv.writer(f)
        writer.writerows(sensor_data[j])

我的sensor_data[j]像这样：

gyro.append((float(merge.iloc[j, 2]), float(merge.iloc[j, 3]), float(merge.iloc[j, 4]))) if merge.iloc[j, 0] == 'Gyroscope' \
             else acc.append((float(merge.iloc[j, 2]), float(merge.iloc[j, 3]), float(merge.iloc[j, 4])))

为什么不打印浮点元组而不是字符串元组？

merge = pd.read_csv(final_user_study_path + "/P" + str(i) + "/DataCollection/data/merge.csv")

合并看起来像：

Gyroscope,50958,-0.3387541,0.014913702,0.05965481
Gyroscope,50978,-0.23009712,0.03621899,0.044741105
LinearAcceleration,50979,-0.09891046,0.119261175,-0.108936176
Gyroscope,50998,-0.1768339,0.066046394,0.040480047
LinearAcceleration,50999,-0.0863409,0.18240824,-0.07152677

所有这些值的类型都是字符串。

我想要：

(-0.16192019, -0.007456851, 0.08841695),(-0.10439591, 0.024501082, 0.12783173),(-0.05965481, 0.049002163, 0.12144014),...

Answer 1

若要停止csv.writer()在需要时添加引号（因为它在单元格中发现了定界符），请添加quoting参数，如下所示：

writer = csv.writer(f, quoting=csv.QUOTE_NONE)

请参阅CSV文档： csv.QUOTE_NONE

但是，最好拆开元组的包装，以便它们各自正确地获得自己的单元格而没有任何多余的括号。 这可以使用itertools.chain.from_iterable() ，将所有元组折叠到一个列表中。 例如：

from itertools import chain

row = [(-0.16192019, -0.007456851, 0.08841695), (-0.10439591, 0.024501082, 0.12783173), (-0.05965481, 0.049002163, 0.12144014)]

print row
print list(chain.from_iterable(row))

这将显示：

[(-0.16192019, -0.007456851, 0.08841695), (-0.10439591, 0.024501082, 0.12783173), (-0.05965481, 0.049002163, 0.12144014)]
[-0.16192019, -0.007456851, 0.08841695, -0.10439591, 0.024501082, 0.12783173, -0.05965481, 0.049002163, 0.12144014]

如您所见，第二行显示所有元组均已展平为一个列表。