python

Anaconda

使用anaconda 安装python包的路径为

~/anaconda2/lib/python2.7/site-packages/

使用 sudo pip命令安装包的目录为：

/usr/local/lib/python2.7/dist-packages/
Official website: http://anaconda.org/

install packages steps

anaconda search -t conda imagemagic
anaconda show kalefranz/imagemagick # 与在网页端搜索 imagemagick 一样

Using conda: http://conda.pydata.org/docs/using/using.html
安装anaconda后需要source ~/.bashrc

基类

class A(object):
  def __init__(self):
    pass

  def get_output_shape_for(self):
    """
    write something
    """
    raise NotImplementedError

slice 切片

>>> slice
<type 'slice'>
>>> slice(None)
slice(None, None, None) # start, stop, step

>>> ls = range(10)
>>> print ls[slice(0, 10, 2)]
[0, 2, 4, 6, 8]
>>>print ls[slice(None)]
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

zip(*)

>>> ls = [(1, 2, 3), (4, 5, 6), (7, 8, 9)]
>>> for s in zip(ls):
      print s
((1, 2, 3), )
((4, 5, 6), )
((7, 8, 9), )
>>> for s in zip(*ls):
      print s
(1, 4, 7)
(2, 5, 8)
(3, 6, 9)

generator

>>> print ((s for s in range(3)))
<type 'generator'>

装饰器

class A(object): ##必须有object
  def __init__(self):
      self.name = 'john'
  @property
  def name(self):
      print 'property'
      return self._name

  @name.setter
  def name(self, name):
      print 'setter'
      self._name = name

>>> a = A()
setter
property

python 处理文件

os.path 模块博客

python built-in functions

os.path offical

shutil.copy("oldfile","newfile")  ＃python 移动文件，oldfile只能是文件夹，newfile可以是文件，也可以是目标目录
os.path.join(name1, name2)
os.path.isdir(dir_name)
os.listdir(directory)
file_name.endswith(suffix, start, end)
os.path.dirname(path) #返回文件路径

python 简单的数据操作

s = '123+'
s[:s.find('+')] == '123'
dic = {}
dic.setdefault('a', []) # dic.get('a', []) and also set dic['a']=[] if 'a' not in dic
print __file__

from numpy.matlib import repmat
repmat(x, m,n) #类似于matlab中的repmat函数

iterator vs generator

blog-using iterators and generators

next(generator_name)

The transformation of images is not under thread lock so it can be done in parallel* 受线程保护的程序不能够并行，
python 互相导入

test1.py, test2.py 两个文件可以互相导入
line_profiler

line_profiler is a module for doing line-by-line profiling of functions. kernprof is a convenient script for running either line_profiler or the Python standard library’s cProfile or profile modules, depending on what is available.
记忆体、递归

@functools.lru_cache(maxsize=None) wiki

python class

class A():
  '''
  class A description
  '''
  def __init__(self):
          pass
  x = 10

>>> A.__dict__
{'x': 10, '__module__': '__main__', '__doc__': '\n\tclass A description\n\t', '__init__': <function __init__ at 0x7f5335e46488>}
>>> A.__doc__
'\n\tclass A description\n\t'

blog-super-considered-super by Deep Thoughts by Raymond Hettinger

super() is in the business of delegating method calls to some class in the instance’s ancestor tree

Things to Know About Python Super

What does super do in Python?-stackoverflow

下面两段代码在 blocks_dict 很大时的速率相差很大，后者明显优于前者

blocks_list_1 = []
for ref in refcode_list:
  if ref in blocks_dict.keys(): # blocks_dict a dictionary
      tmp = blokc_dict[ref]
      blocks_list_1.append(tmp)
  else:
      continue

blocks_list_1 = []
for ref in refcode_list:
  try:
      tmp = blokc_dict[ref]
      blocks_list_1.append(tmp)
  except Exception:
      continue

切换python运行版本

ls /usr/bin/python* python-versions

vim ~/.bashrc

# add this line in the file bottom
alias python='/usr/bin/python2.7'
. ~/.bashrc

python 优雅的操作字典

https://www.linuxzen.com/python-you-ya-de-cao-zuo-zi-dian.html
Numpy 读.txt文件，生成矩阵

http://hyry.dip.jp/tech/book/page/scipy/numpy_file.html
美团点评技术团队

http://tech.meituan.com/ http://www.w3ctech.com/

Anaconda

The transformation of images is not under thread lock so it can be done in parallel* 受线程保护的程序不能够并行，