1. 列表
#以下三式等价 c = (a>b and a or b) c = a if a>b else b c = [b, a][a>b]
字符串拼接
' + '.join('%s,%s'%(a,b) for a,b in zip(list('abc'), range(3)))
'a,0 + b,1 + c,2'
字符查找
'aa' in '大家aa' #True
'大家aa'.find('大家') #0
for循环
from itertools import product result = product([1,2,3], ['a','b']) #[(1, 'a'), (1, 'b'), (2, 'a'), (2, 'b'), (3, 'a'), (3, 'b')]
数组重构，np.newaxis
np.arange(5)[:,None]
array([[0],
[1],
[2],
[3],
[4]])
数组拼接， np.r_[], np.c_[]，注意是方括号
np.r_[np.array([1,2,3]), np.array([1,2,3])] #array([1, 2, 3, 1, 2, 3]) np.c_[np.array([1,2,3]), np.array([1,2,3])] #array([[1, 1], # [2, 2], # [3, 3]]) np.r_[np.array([[1,2,3],[1,2,3]]), np.array([[4,5,6],[4,5,6]])] #array([[1, 2, 3], # [1, 2, 3], # [4, 5, 6], # [4, 5, 6]]) np.c_[np.array([[1,2,3],[1,2,3]]), np.array([[4,5,6],[4,5,6]])] #array([[1, 2, 3, 4, 5, 6], # [1, 2, 3, 4, 5, 6]])
2. 嵌套列表展开，嵌套方法可实现多层列表展开
a=[[1,2],[3,4],[5,6]]
print([j for i in a for j in i]) #列表生成式
sum(a, []) #求和函数
from functools import reduce
reduce(lambda x,y: x+ y, a) #工具函数
from itertools import chain
print(list(chain(*a)))
import itertools
print(list(itertools.chain.from_iterable(a)))
t=[]
[t.extend(i) for i in a]
print(t)
#递归的方法比较容易理解
#在stackoverflow看到大牛的列表生成式版本
func = lambda x: [y for l in x for y in func(l)] if type(x) is list else [x]
#递归自定义
def expand_list(nested_list):
for item in nested_list:
if isinstance(item, (list, tuple)):
for sub_item in expand_list(item):
yield sub_item
else:
yield item
3. 工具函数
# map 把函数 f 依次作用在 list 的每个元素上，得到一个 iterator 并返回。 print(list(map(format_name, ['adam', 'LISA', 'barT']))) # reduce()传入的函数 f 必须接收两个参数，reduce()对list的每个元素反复调用函数f，并返回最终结果值。reduce()还可以接收第3个可选参数，作为计算的初始值。 from functools import reduce print(reduce(f, [1, 3, 5, 7, 9], 100)) # filter()根据判断结果自动过滤掉不符合条件的元素，返回由符合条件元素组成的iterator。 print(list(filter(is_odd, [1, 4, 6, 7, 9, 12, 17]))) #zip()参数可以接受任何类型的序列，同时也可以有两个以上的参数;当传入参数的长度不同时，zip能自动以最短序列长度为准进行截取，获得元组 [i for i in zip(['a','b'], (3,4,5))]
4. 时间
Python处理时间 time && datetime 模块
import time
time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(time.time()))
5. 随机数
import numpy as np np.random.rand(10) np.random.randn(10) np.random.randint(10, size=10) np.random.shuffle(x) #类似洗牌，打乱顺序 np.random.permutation(citys.index) #返回一个随机排列 import random random.random() #只产生一个 random.randint(1, 10) random.uniform() random.randrange() #[random.randrange(0,9) for i in range(30)] random.shuffle(a) #a=[1,2,3], 直接对a的值修改
6. dataframe迭代（iteritems, iterrows, itertuples, apply）
import pandas as pd
df = pd.DataFrame([{'c1':10, 'c2':100}, {'c1':11,'c2':110}, {'c1':12,'c2':120}])
[(i,j) for i,j in df.T.iteritems()]
[(i,j) for i,j in df.iterrows()]
[(i, j['c1'], j['c2']) for i,j in df.iterrows()]
[(getattr(row, "c1"), getattr(row, "c2")) for row in df.itertuples(index=True, name='Pandas')]
def valuation_formula(x, y):
return x * y * 0.5
df['price'] = df.apply(lambda row: valuation_formula(row['c1'], row['c2']), axis=1)
7. dataframe修改索引
a.rename(columns={'A':'a', 'C':'c'}, inplace = True) #dataframe修改列名
a.index = ['1', '2', np.nan] # dataframe修改索引
8. 读写文件
读文件
import pandas as pd
import os
excel_data = pd.read_excel("./xx/xx.xlsx")
csv_data = pd.read_csv("./xx/xx.csv")
txt_data = pd.read_table("./xx/xx.txt")
json_data = pd.read_json("./xx/xx.json")
df = pd.read_clipboard(header=None) #从剪切板读取文件，参数同 pd.read_table
print('excel数据', os.linesep, excel_data)
分组txt文件读写
def write_list(path, li):
extension = path.split('.')[-1]
if extension == 'txt':
with open(path, 'w', encoding='utf-8') as fp:
#fp.writelines(li)
fp.write('
'.join(li))
elif extension == 'csv':
pd.DataFrame({'x':li}).to_csv(path, index=False, header=False)
else:
raise ValueError('扩展名只能指定为txt或csv...')
def write_by_group(id_dict, outpath, size, extension):
import os
import math
import pandas as pd
if '/' in outpath:
outfile = outpath + '/{}_{}.' + extension
else:
outfile = os.path.join(outpath, '{}_{}.txt')
for gro_ in id_dict.keys():
ids = id_dict[gro_]
st, ed = 0, 0
for i in range(math.ceil(len(ids)/size)):
ed += size
write_list(outfile.format(gro_, i), ids[st:ed])
st = ed
def main(infile, outfile, size, extension):
from collections import defaultdict
with open(infile, 'r', encoding='utf-8') as fp:
id_dict = defaultdict(lambda: [])
for s in fp.readlines():
id_,gro_ = s.strip('
').split('|')
id_dict[gro_] += [id_]
write_by_group(id_dict, outpath, size, extension)
if __name__ == "__main__":
infile = r'e:新建文本文档.txt'
outpath = 'e:/新建文件夹'
size = 3
main(infile, outpath, size, 'txt')
txt文件读写
import math
def write_list(path, li):
with open(path, 'w', encoding='utf-8') as fp:
fp.writelines(li)
def main():
with open('aa.txt', 'r', encoding='utf-8') as fp:
ids = [s.split('|')[0]+'
' for s in fp.readlines()]
n = 10
size = math.ceil(len(ids)/n)
st, ed = 0, 0
for i in range(n):
ed += size
write_list('e:a{}.txt'.format(i), ids[st:ed])
st = ed
if __name__ == "__main__":
main()
内存文件读写
from io import StringIO
#f = StringIO()
#f.write('Hello World!')
f = StringIO('Hello World!')
string = f.readlines()
print(string)
批量读取大文件
'''
报错：OSError: Initializing from file failed
解决：pd.read_csv('xxx.csv', chunksize=100, engine='python')
'''
df = pd.read_csv('xxx.csv', chunksize=1000000)
for chunk in df: print(chunk.shape)
df = pd.read_csv('xxx.csv', iterator=True)
df.get_chunk(1000000).shape
def get_df(file):
mylist = []
for chunk in pd.read_csv(file, chunksize=1000000):
mylist.append(chunk)
temp_df = pd.concat(mylist, axis=0)
del mylist
return temp_df.shape
写文件
df.to_csv("./xx/xx.csv", index=False, encoding='utf-8')
df.to_excel("./xx/xx.xlsx") #utf_8_sig, gbk, gb2312, gb18030
df.to_dict(orient = 'records') #orient默认dict（字典套字典）, 还可 ‘list’, ‘series’, ‘split’, ‘records’, ‘index’
excel文件的读写
import xlrd
file_path = r'F:/test.xlsx' #路径前加 r，读取的文件路径
file_path = file_path.decode('utf-8') #文件路径的中文转码
data = xlrd.open_workbook(file_path) #获取数据
table = data.sheet_by_name('Sheet1') #获取sheet
nrows = table.nrows #获取总行数
ncols = table.ncols #获取总列数
rowvalue = table.row_values(5) #获取一行的数值，例如第5行
col_values = table.col_values(6) #获取一列的数值，例如第6列
cell_value = table.cell(5,6).value #获取一个单元格的数值，例如第5行第6列
excel文件读写
import xlwt
import xlrd
from xlutils.copy import copy
#打开需要操作的excel表
wb=xlrd.open_workbook(path)
#复制原有表
newb=copy(wb)
#新增sheet,参数是该sheet的名字，可自定义
wbsheet=newb.add_sheet(dl+'-'+dn)
#向新sheet中写入数据。本代码中的d是某个dataframe
wbsheet.write(0,0,'date')
wbsheet.write(0,1,'visited')
wbsheet.write(0,2,'success')
for i in range(d.shape[0]):
wbsheet.write(i + 1, 0, d.iloc[i, 0])
for j in range(1,d.shape[1]):
wbsheet.write(i+1,j,int(d.iloc[i,j]))
#获取原有excel表中sheet名为‘summary'的sheet
sumsheet=newb.get_sheet('summary')
#k表示该sheet的最后一行
k=len(sumsheet.rows)
#想原有sheet后面新增数据
sumsheet.write(k,0,dl+'-'+dn)
sumsheet.write(k,1,int(sum(d['visited'])))
sumsheet.write(k,2,int(sum(d['success'])))
#保存为原有的excel表路径
newb.save(path)
pd.read_excel，pd.to_excel参数
pandas.read_excel(io, sheet_name=0, header=0, skiprows=None, skip_footer=0, index_col=None, names=None, usecols=None, parse_dates=False, date_parser=None, na_values=None, thousands=None, convert_float=True, converters=None, dtype=None, true_values=None, false_values=None, engine=None, squeeze=False, **kwds) DataFrame.to_excel(excel_writer, sheet_name='Sheet1', na_rep='', float_format=None, columns=None, header=True, index=True, index_label=None, startrow=0, startcol=0, engine=None, merge_cells=True, encoding=None, inf_rep='inf', verbose=True, freeze_panes=None)
追加excel文件，不覆盖原有数据
def excelAddSheet(dataframe, filepath, sheet_name):
import os
from openpyxl import load_workbook
if os.path.exists(filepath):
#当表名已存在时，后面还可以添加参数，进行追加
book = load_workbook(filepath) #FileNotFoundError
excelWriter = pd.ExcelWriter(filepath, engine='openpyxl')
excelWriter.book = book
dataframe.to_excel(excel_writer=excelWriter, sheet_name=sheet_name, index=None)
excelWriter.save()
excelWriter.close()
else:
dataframe.to_excel(filepath, sheet_name=sheet_name, index=None)
excelAddSheet(df2[:2], filepath, 'Sheet1')
excelAddSheet(df2[:2], filepath, 'Sheet2')
9. 数据库文件读写
import pandas as pd
from sqlalchemy import create_engine, types
config = {db_flag : "mysql", host_ip : "127.0.0.1", host_port : 3306, db_name : "test", table_name : "aaa", user : "root", pawd : "root", charset : "utf-8"}
engine_config = '''mysql+pymysql://{user}:{pawd}@{host_ip}:{host_port}/{db_name}?charset={charset}'''.format(**config)
engine= create_engine(engine_config)
pd.read_sql_query(sql, engine)
conn = engine.connect()
df.to_sql(table_name, conn, if_exists='append', index=False)
10. 断言 - 单元测试
assert 1==1
assert 1==2
#Traceback (most recent call last):
# File "<ipython-input-174-730332727407>", line 1, in <module>
# assert 1==2
#AssertionError
#等价于
if not 1==2:
raise AssertionError
#Traceback (most recent call last):
# File "<ipython-input-176-bd2dcc9e33c8>", line 2, in <module>
# raise AssertionError
#AssertionError
#try ... except .. else
#自定义新异常
class Networkerror(RuntimeError):
def __init__(self, arg):
self.args = arg
try:
raise Networkerror("Bad hostname")
except Networkerror,e:
print e.args