这些类型之间有什么区别 <type 'numpy.string_'> 和 <type 'str'> ？

Question

类型<type 'numpy.string_'>和<type 'str'> <type 'numpy.string_'>之间是否存在差异？

Answer 1

numpy.string_ is the NumPy datatype used for arrays containing fixed-width byte strings. On the other hand, str is a native Python type and can not be used as a datatype for NumPy arrays*.

If you create a NumPy array containing strings, the array will use the numpy.string_ type (or the numpy.unicode_ type in Python 3). More precisely, the array will use a sub-datatype of np.string_ :

>>> a = np.array(['abc', 'xy'])
>>> a
array(['abc', 'xy'], dtype='<S3')
>>> np.issubdtype('<S3', np.string_)
True

In this case the datatype is '<S3' : the < denotes the byte-order (little-endian), S denotes the string type and 3 indicates that each value in the array holds up to three characters (or bytes).

One property that np.string_ and str share is immutability. Trying to increase the length of a Python str object will create a new object in memory. Similarly, if you want fixed-width NumPy array to hold more characters, a new larger array will have to be created in memory.

---

* Note that it is possible to create a NumPy object array which contains references to Python str objects, but such arrays behave quite differently to normal arrays.