Python 从HDFStore目录错误中恢复

我制作了一个新的HDFStore并在里面放了一些东西

In [1]: store = pd.HDFStore('data.h5')

In [2]: store
Out[2]: 
<class 'pandas.io.pytables.HDFStore'>
File path: data.h5
Empty

In [3]: store['a'] = Series([1, 2])

In [4]: store
Out[4]: 
<class 'pandas.io.pytables.HDFStore'>
File path: data.h5
/a            series       (shape->[2])

In [5]: store.close()

In [6]: exit

哎呀！意识到我的错误，我将

cd

放入正确的目录，并在正确的目录中重新分配

存储

变量：

In [3]: cd
/home/dallan

In [4]: store = pd.HDFStore('data.h5')

In [5]: store
Out[5]: 
<class 'pandas.io.pytables.HDFStore'>
File path: data.h5
Empty

In [6]: exit

按照杰夫的建议，我尝试

！cd

显示了意外的IPython行为：

In [1]: pwd
Out[1]: u'/home/dallan'

In [2]: !cd some_directory/

In [3]: pwd
Out[3]: u'/home/dallan'

In [4]: !cd /home/dallan/some_directory/

In [5]: pwd
Out[5]: u'/home/dallan'

In [6]: cd some_directory/
/home/dallan/some_directory

In [7]: pwd
Out[7]: u'/home/dallan/some_directory'

---

它直接在“当前”中查看，我不确定ipython中的

cd

是否确实改变了它。实际上，您正在查看另一个文件。尝试绝对路径，或者

！cd

这可能是HDFStore中的错误吗？（你能用to_csv/read_csv做这件事吗？）。为什么除了HDFStore之外，

cd

什么都可以用？

In [1]: store = pd.HDFStore('data.h5')

In [2]: store
Out[2]: 
<class 'pandas.io.pytables.HDFStore'>
File path: data.h5
/a            series       (shape->[2])

In [1]: Series([1, 2]).to_csv('test.txt')

In [2]: exit

In [1]: cd some_directory/
/home/dallan/some_directory

In [2]: s = pd.read_csv('test.txt')
(...) IOError: File test.txt does not exist

In [3]: cd
/home/dallan

In [4]: s = pd.read_csv('test.txt')

In [5]: s
Out[5]: 
   0  1
0  1  2

In [1]: pwd
Out[1]: u'/home/dallan'

In [2]: !cd some_directory/

In [3]: pwd
Out[3]: u'/home/dallan'

In [4]: !cd /home/dallan/some_directory/

In [5]: pwd
Out[5]: u'/home/dallan'

In [6]: cd some_directory/
/home/dallan/some_directory

In [7]: pwd
Out[7]: u'/home/dallan/some_directory'