import requests

ret = requests.get('https://api.day.app/8BZtwxVav***********/报警滴滴滴/测试消息')

import requests
import json

# wxpusher
headers = {'content-type': "application/json"}
body = {
  "appToken":"AT_yn7Xsvz**********",
  "content":"这是一条测试消息",
  "summary":"报警滴滴滴",
  "contentType":1,
  "topicIds":[],
  "uids":["UID_VO8eFt***********"]
}

ret = requests.post('http://wxpusher.zjiecode.com/api/send/message', data=json.dumps(body), headers=headers)

cols = ['col1', 'col2']
pd.read_table(path, sep='|', headers=None, names=cols)
pd.head()	# first five rows

df.describe()	#show mean, std, max, min...

df.rename(columns = {'old_col_name':'new_col_name'}, inplace=True)

cols = new_col_list
df.columns = cols
df.columns = df.columns.str.replace(' ', '_')

df.drop('col', axis=1, inplace=True)

df.drop([rows], axis=0, inplace=True)

df['col'].sort_values(ascending=False)	# series

df.sort_values('col')	# return dataframe
df.sort_values(['col1', 'col2'])	

pd.Series([])

df[df['col'] >= 200]

#	df[df['col'] >= 200]['col2']
df.loc[df['col'] >= 200, 'col2']	#	loc select cols or rows with labels

df[(df['col'] >= 200) | (df['col1'] == 'a')]

df['col'].isin(['a', 'b', 'c'])

pd.read_csv(path, usecols=['a', 'b'])

pd.read_csv(path, nrows=3)

for c in df['col']:
	print(c)

for index, row in df.iterrows():
  print(index, row['a'], row['b'])

import numpy as np
df.select_dtypes(include=[np.number]).dtypes

df.describe(include=['object', 'float64'])

df.mean()	# mean value for serires
df.mean(axis=1)	# row mean

df['col'].str.upper()	# uppercase 
df['col'].str.contains('str')	#	return series of t/f
df[df['col'].str.contains('str')]
df['col'].str.replace('a', 'b').str.replace('c', 'd')	# support regex

df.dtypes
serires.dtype
df['col'] = df['col'].astype('float64')

df = pd.read_csv(path, dtype=('col1':float))
df['col'].str.replace('$', '').astype(float).mean()

df.groupby('col1').mean()
df.groupby('col1')['col2'].mean()
df.groupby('col1')['col2'].agg(['count', 'min', 'max', 'mean'])

%matplotlib inline
df.groupby('col1')['col2'].mean().plot(kind='bar')

df['col'].describe()
df['col'].value_counts()
df['col'].value_counts(normalize=True)
df['col'].unique()
df['col'].nunique()	# number of unique values
pd.crosstab(df['col1'], df['col2'])
df['col1'].plot(kind='hist')

df.tail()
df.isnull().head()
df.isnull().sum()
df.dropna(how='any', inplace=True)	# drop a row if it contains null
df.dropna(subset=['col1', 'col2'], how='any')	# drop a row if col1 or col2 is null
df['col'].value_counts(dropna=False)
df['col'].fillna(value='A', inplace=True)

df.index
df.columns
df.shape

df.loc[23, 'a']	#	fetch value
df.decribe().loc['25%', 'col']

df.set_index(Series, inplace-True)

df['col'].value_counts()['value1']
df['col'].value_counts().sort_values()
df['col'].value_counts().sort_index()

pd.Series(value array, index=[indices])
pd.concat([df1, df2], axis=1)

df.loc[index, col_name]	# by label
df.iloc[:, [0, 3]]	# by indices

df= pd.read_csv(path, index_col='col1')
df.ix['row_label', 0]	# mix labels and indices (not recommend)
df.ix[1, 'col_label']
df.ix['col1':'col3', 0:2]

df.memory_usage(deep=True)
df['col'].astype('category')
df['col'] = df['col'].astype('category', categories=['good', 'veray good', 'excellent', ordered=True])

pd.DataFrame({'id':array1, 'col':array2}).set_index('id').to_csv('1.csv')
df.to_pickle('1.pkl')
df.read_pickle('1.pkl')

df.loc[0:4, :]	#	inclusive-label
df.iloc[0:4, :]	# exclusive-index

df.sample(n=3, random_state=42)
df.sample(frac=0.75)
train = df.sample()
test = df.loc[~df.index.isin(train.index), :]

df['sex_male'] = df['sex'].map({'female':0, 'male':1})
pd.get_dummies(df['sex'], prefix='sex')

df['time'] = pd.to_datetime(df['time'])
df['time'].dt.weekday_name
df['time'].dt.dayofyear
df.loc[df['time'] > pd.to_datetime('1/1/1999'), :]
(df['time'].max() - df['time'].min()).days

df['col'].duplicated().sum()
df.loc[df.duplicated(keep='first'), :]	#	'last' False
df.duplicated(subset=['col1', 'col2'])

pd.set_option('display.max_rows', None)
pd.reset_option('display.max_rows', None)
pd.set_option('display.max_colwidth, None)
pd.set_option('display.precision, 2)
df['new_col'] = df['col'] * 1000
pd.set_option('display.float_format', '{:,}'.format)

pd.DataFrame(dict, columns=['a', 'b'])
s = pd.Series([1, 2], index=['c', 'd'], name=shape)

df['col1'] = df['col2'].apply(function)
df['col1'] = df['col2'].apply(np.ceil)
df['col'].apply(lambda x: x[0])
df['col1'] = df['col2'].apply(np.argmax, axis=1)

ser = df.groupby(['col1', 'col2']).col3.mean()	# series with multiindex
ser.unstack()	# dataframe
df.pivot_table(values='col1', index='col2', columns='col3')

df.set_index(['col1', 'col2'], inplace=True)
df.sort_index(inplace=True)

pd.concat([df1, df2])
pd.merge([df1, df2])

pd.merge(df1, df2, left_on='col1', right_on='col2')
pd.merge(df1, df2, how='inner')	#	outer left right

变量名 赋值符号 变量名

id()

type()

print('-' * 10)
# -----------

list2 = list1.copy()

import copy
list3 = copy.deepcopy(list1)

a = {
  "key":"value"
	"1":1,
  "2":"qwe"
}

del x	# 解除变量与值的绑定

input() # 存为str

# python2
raw_input()	# 存为str
input()	# 要求用户输入明确数据类型，输入的是什么类型就为什么类型

print("name: %(name)s, age:%(age)d" % {"name": "a", "age": 10})
# %s可以接受任何类型

print("name: {name}, age:{age}".format(name = "a", age = 10))

print(f"name: {name}, age:{age}")

10 // 3	#只保留整数部分，即整除
3 ** 10	#指数
10 % 3 #取余数

1 != 2
1 < x < 3

a = 1

 a += 1
 b *= 1

z = y = x = 10 

m, n = n, m

list = [1, 2, 3, 4]
a, b, c, d = list
x, y, *_ = list	#取前两个
*_ , x, y = list	#取后两个
# 字典解压取值的是key

"a" in "abc"
1 in [1, 2, 3]

if 16 < a < 20:

while True:
	...
else:
	...

for variable in Iterable object:
	...
  

for i in range(1, 100):
	...
  

# py2
>>> range(0, 5)
[0, 1, 2, 3, 4]
#py3 优化
>>> range(0,5)
range(0,5)

print("hello", end="*")	# 自定义结束，默认为换行

# 不可变，是一个整体，不可单独改某个字符
str(10)
msg = "hello"

# 切片
msg[0] # h
msg[-1] # o
msg[0:5] # hello 顾头不顾尾
msg[0:5:2] # hlo
msg[5:0:-1] # olle

"alex" in ”alexxxx"	# true
"alex" not in ”alexxxx"	# false

msg = "    eee   "
res = msg.strip()	# 默认去除两边空格 字符串不可变，故需要新赋值
msg = "***eee***"
res = msg.strip(“*”)  # strip只去两边，不去中间
res = msg.lstrip()
res = msg.rstrip()

# 切分：按某字符分割字符串，返回列表
res = msg.split(":", 1) # 默认按空格为分隔符 分割次数，从左往右算
res = msg.lsplit(":", 1)
res = msg.rsplit(":", 1) #从右往左

res2 = ":".join(res) # 以：为分隔符拼接

res = msg.lower() # 小写
res = msg.upper()
res = msg.startswith("aaa") # True 以什么开头
res = msg.endswith("aaa")
res = msg.replace("you", "me" ,1) #字符串替换，次数

“123”.isdigit() # 纯数字

msg.find("e") # 返回索引，找不到返回-1
msg.index("e") # 返回索引，找不到程序报错
msg.count("e") # 数子串出现次数
"分割线".center(50, "*") #两侧用*填充
"分割线".ljust(50, "*") #右侧填充
"分割线".rjust(50, "*")
"分割线".zfill(50) # 左侧用0填充

# is 判断系列 查文档
isalpha()
isdigit()
isdecimal() # 只识别阿拉伯数字 py3 默认unicode
isnumeric() # 可以识别汉字，阿拉伯数字