1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
|
Metadata-Version: 1.1
Name: feather-format
Version: 0.3.1
Summary: Python interface to the Apache Arrow-based Feather File Format
Home-page: http://github.com/wesm/feather
Author: Wes McKinney
Author-email: wesm@apache.org
License: Apache License, Version 2.0
Description: ## Python interface to the Apache Arrow-based Feather File Format
Feather efficiently stores pandas DataFrame objects on disk.
## Installing
```shell
pip install feather-format
```
From [conda-forge][1]:
```shell
conda install feather-format -c conda-forge
```
#### Mac notes
Anaconda uses a default 10.5 deployment target which does not have C++11
properly available. This can be fixed by setting:
```
export MACOSX_DEPLOYMENT_TARGET=10.9
```
Deployments targets as early as 10.7 can be used if the compiler supports
C++11 and the correct mode is selected. For example using the following:
```
export MACOSX_DEPLOYMENT_TARGET=10.7
export CFLAGS="${CXXFLAGS} -stdlib=libc++ -std=c++11"
export CXXFLAGS="${CXXFLAGS} -stdlib=libc++ -std=c++11"
```
This may be necessary in some other OS X environments.
## Build
Building Feather requires a C++11 compiler. We've simplified the PyPI packaging
to include libfeather (the C++ core library) to be built statically as part of
the Python extension build, but this may change in the future.
### Static builds for easier packaging
At the moment, the libfeather sources are being built and linked with the
Cython extension, rather than building the `libfeather` shared library and
linking to that.
While we continue to do this, building from source requires you to symlink (or
copy) the C++ sources. See:
```shell
# Symlink the C++ library for the static build
ln -s ../cpp/src src
python setup.py build
# To install it locally
python setup.py install
# Source distribution
python setup.py sdist
```
To change this and instead link to an installed `libfeather.so`, look in
`setup.py` and make the following change:
```python
FEATHER_STATIC_BUILD = False
```
## Limitations
Some features of pandas are not supported in Feather:
* Non-string column names
* Row indexes
* Object-type columns with non-homogeneous data
[1]: https://conda-forge.github.io
Platform: UNKNOWN
Classifier: Development Status :: 3 - Alpha
Classifier: Environment :: Console
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3.4
Classifier: Programming Language :: Python :: 3.5
Classifier: Programming Language :: Cython
|
