python实战之实现excel读取、统计、写入
背景
图像领域内的一个国内会议快要召开了，要发各种邀请邮件，之后要录入、统计邮件回复（参会还是不参会等）。如此重要的任务，老师就托付给我了。ps: 统计回复邮件的时候，能知道谁参会或谁不参会。
而我主要的任务，除了录入邮件回复，就是统计理事和普通会员的参会情况了（参会的、不参会的、没回复的）。录入邮件回复信息没办法只能人工操作，但如果统计也要人工的话，那工作量就太大了（比如在上百人的列表中搜索另外上百人在不在此列表中！！），于是就想到了用python来帮忙，花两天时间不断修改，写了6个版本。。。
摘要
- version_1基本实现了
excel读取、统计、显示功能，但问题也有不少，像显示出来后还要自已复制、粘贴到excel表，而且set中还有nan这样的bug。 version_2相比较version_1而言，此版本用set代替list，可以自动去重。version_3解决了set中出现nan的bug，而且还加入的excel写入的功能，但一次只能写入一张表，所以要运行两次才能写入两张表（sheet）。version_4的改进在于将version_3中写入两张表格的操作，集成在一个程序里，只需要运行一次便可写入两张表，但也总是会写入两张表，万一你只想写入一张表呢？？version_5相对之前版本的最大改进在于将程序模块化，更具可读性了; 对修复set中出现nan的方法也进行了改进和简化; 而且可以自由控制写入多少张表了。version_final相比较version_5，修复了一个bug，之前需要先验知识，现在更通用一点（prep函数取代了set2list函数）。
version_1
基本实现了excel读取、统计、显示功能，但问题也有不少，像显示出来后还要自已复制、粘贴到excel表，而且set中还有nan这样的值。
#version_1
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop\\0711任务')
print(os.getcwd())
data = pd.read_excel('for_python.xlsx','Sheet2')
return_set = set(data['回执名单'])
demand_set = set(data['理事名单'])
answer_list = []
unanswer_list = []
for each in demand_set:
if each in return_set:
answer_list.append(each)
else:
unanswer_list.append(each)
notattend_set = set(data['回执名单'][-15:])
nt = []
for each in notattend_set:
if each in answer_list:
nt.append(each)
def disp(ll, cap, num = True):
print(cap)
if num:
for i, each in enumerate(ll):
print(i+1,each)
else:
for each in enumerate(ll):
print(each)
disp(answer_list,'\n理事回执名单')
disp(unanswer_list,'\n理事未回执名单')
disp(nt,'\n理事回执说不参加名单')
version_2
相比较上一个版本，此版本用set代替list，可以自动去重。
#version_2
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop\\0711任务')
print(os.getcwd())
data = pd.read_excel('for_python.xlsx','Sheet2')
return_set = set(data['回执名单'])
demand_set = set(data['理事名单'])
answer_set = set([]) #理事回执名单
unanswer_set = set([]) #理事未回执名单
for each in demand_set:
if each in return_set:
answer_set.add(each)
else:
unanswer_set.add(each)
notattend_set = set(data['回执名单'][-17:])
nt = set([]) #理事回执说不参加名单
for each in notattend_set:
if each in answer_set:
nt.add(each)
ans_att_set = answer_set - nt #理事回执参加名单
def disp(ss, cap, num = False):
print(cap)
if num:
for i, each in enumerate(ss):
print(i+1,each)
else:
for each in ss:
print(each)
#disp(answer_set,'\n理事回执名单')
disp(ans_att_set,'\n理事回执说参加名单')
disp(nt,'\n理事回执说不参加名单')
disp(unanswer_set,'\n理事未回执名单')
print(len(ans_att_set),len(nt),len(unanswer_set))
version_3
此版本解决了set中出现nan的bug，而且还加入的excel写入的功能，但一次只能写入一张表，所以要运行两次才能写入两张表（sheet）。
step_1
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop')
print('work_directory: ', os.getcwd())
data = pd.read_excel('理事与会员名单.xlsx','理事与会员名单')
#1.载入excel，得到三个名单
ans_attend_set = set(data['回执参加']) #回执参会名单
N = len(ans_attend_set)
ans_notatt_idx = [i for i in range(N) if type(data['回执不参加'][i]) == np.float][0]
ans_notatt_set = set(data['回执不参加'][:ans_notatt_idx])#回执不参会名单
concil_idx = [i for i in range(N) if type(data['理事名单'][i]) == np.float][0]
concil_set = set(data['理事名单'][:concil_idx]) #理事名单
#2.统计理事参会情况
concil_attend_set = set([]) #理事回执参会名单
concil_notatt_set = set([]) #理事回执不参会名单
concil_notans_set = set([]) #理事未回执名单
for each in concil_set:
if each in ans_attend_set:
concil_attend_set.add(each)
elif each in ans_notatt_set:
concil_notatt_set.add(each)
else:
concil_notans_set.add(each)
#3. 显示结果
def disp(ss, cap, num = True):
#ss: 名单集合
#cap: 开头描述
print(cap,'({})'.format(len(ss)))
for i in range(np.ceil(len(ss)/5).astype(int)):
pre = i * 5
nex = (i+1) * 5
#调整显示格式
dd = ''
for each in list(ss)[pre:nex]:
if len(each) == 2:
dd = dd + ' ' + each
elif len(each) == 3:
dd = dd + ' ' + each
else:
dd = dd + '' + each
print('{:3.0f} -{:3.0f} {}'.format(i*5+1,(i+1)*5,dd))
disp(concil_attend_set,'\n参会理事')
disp(concil_notatt_set,'\n不参会理事')
disp(concil_notans_set,'\n未回执理事')
#4. 将理事参会情况，写入excel
df = pd.DataFrame(list(concil_attend_set),columns = ['参会理事'])
df['']=pd.DataFrame([''])
df['序号1'] = pd.DataFrame(np.arange(len(concil_notatt_set))+1)
df['不参会理事'] = pd.DataFrame(list(concil_notatt_set))
df['_']=pd.DataFrame([''])
df['序号2'] = pd.DataFrame(np.arange(len(concil_notans_set))+1)
df['未回执理事'] = pd.DataFrame(list(concil_notans_set))
df.index = df.index + 1
df.to_excel('理事和会员回执统计.xlsx', sheet_name='理事回执统计')
print('\n\n写入excel成功~~')
step_2
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop')
print('work_directory: ', os.getcwd())
data = pd.read_excel('理事与会员名单.xlsx','理事与会员名单')
#1.载入excel，得到三个名单
ans_attend_set = set(data['回执参加']) #回执参会名单
N = len(ans_attend_set)
ans_notatt_idx = [i for i in range(N) if type(data['回执不参加'][i]) == np.float][0]
ans_notatt_set = set(data['回执不参加'][:ans_notatt_idx])#回执不参会名单
mem_idx = [i for i in range(N) if type(data['被推荐人'][i]) == np.float][0]
mem_set = set(data['被推荐人'][:mem_idx]) #被推荐为会员代表名单
#2.统计会员参会情况
mem_attend_set = set([]) #回执参会会员
mem_notatt_set = set([]) #回执不参会会员
mem_notans_set = set([]) #未回执会员
for each in mem_set:
if each in ans_attend_set:
mem_attend_set.add(each)
elif each in ans_notatt_set:
mem_notatt_set.add(each)
else:
mem_notans_set.add(each)
#3. 显示结果
def disp(ss, cap, num = True):
#ss: 名单集合
#cap: 开头描述
print(cap,'({})'.format(len(ss)))
for i in range(np.ceil(len(ss)/5).astype(int)):
pre = i * 5
nex = (i+1) * 5
#调整显示格式
dd = ''
for each in list(ss)[pre:nex]:
if len(each) == 2:
dd = dd + ' ' + each
elif len(each) == 3:
dd = dd + ' ' + each
else:
dd = dd + '' + each
print('{:3.0f} -{:3.0f} {}'.format(i*5+1,(i+1)*5,dd))
disp(mem_attend_set,'\n参会会员')
disp(mem_notatt_set,'\n不参会会员')
disp(mem_notans_set,'\n未回执会员')
#4. 将会员参会情况，写入excel
if len(mem_attend_set) > len(mem_notans_set):
print('#1')
L = len(mem_attend_set)
mem_notans_list = list(mem_notans_set)
mem_notans_list.extend([''] * (L - len(mem_notans_set)))
mem_attend_list = list(mem_attend_set)
else:
print('#2')
L = len(mem_notans_set)
mem_attend_list = list(mem_attend_set)
mem_attend_list.extend([''] * (L - len(mem_attend_set)))
mem_notans_list = list(mem_notans_set)
df = pd.DataFrame(mem_attend_list,columns = ['参会会员'])
df['']=pd.DataFrame([''])
if len(mem_notatt_set) == 0:
df['序号1'] = np.NaN
df['不参会会员'] = np.NaN
else:
df['序号1'] = pd.DataFrame(np.arange(len(mem_notatt_set))+1)
df['不参会会员'] = pd.DataFrame(list(mem_notatt_set))
df['_']=pd.DataFrame([''])
df['序号2'] = pd.DataFrame(np.arange(len(mem_notans_set))+1)
df['未回执会员'] = pd.DataFrame(mem_notans_list)
df.index = df.index + 1
df0 = pd.read_excel('理事和会员回执统计.xlsx',sheet_name='理事回执统计')
writer = pd.ExcelWriter('理事和会员回执统计.xlsx')
df0.to_excel(writer, sheet_name='理事回执统计')
df.to_excel(writer, sheet_name='会员回执统计')
writer.save()
print('\n\n写入excel成功~~')
version_4
version_4的改进在于将version_3中写入两张表格的操作，集成在一个程序里，只需要运行一次便可写入两张表，也总是会写入两张表。问题是要是你只想写入一张表呢？？
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop')
print('work_directory: ', os.getcwd())
loadfile_sheet = ['理事与会员名单.xlsx','理事与会员名单']
columns = ['回执参加','回执不参加','理事','会员']
savefile_sheet = ['理事和会员回执统计.xlsx','理事回执统计','会员回执统计']
display = [1,1]
def main(loadfile_sheet,columns,savefile_sheet,display):
#1. 载入excel，得到名单
data = pd.read_excel(loadfile_sheet[0],loadfile_sheet[1])
def first_nan_index(pd):
for i, each in enumerate(pd):
if type(each) == np.float:
return i
return i
idx = first_nan_index(data[columns[0]])
ans_attend_set = set(data[columns[0]][:idx])#回执参会名单
idx = first_nan_index(data[columns[1]])
ans_notatt_set = set(data[columns[1]][:idx])#回执不参会名单
idx = first_nan_index(data[columns[2]])
concil_set = set(data[columns[2]][:idx])#理事名单
idx = first_nan_index(data[columns[3]])
mem_set = set(data[columns[3]][:idx])#会员名单
#2. 统计参会情况
concil_attend_set = set([]) #回执参会理事
concil_notatt_set = set([]) #回执不参会理事
concil_notans_set = set([]) #未回执理事
for each in concil_set:
if each in ans_attend_set:
concil_attend_set.add(each)
elif each in ans_notatt_set:
concil_notatt_set.add(each)
else:
concil_notans_set.add(each)
mem_attend_set = set([]) #回执参会会员
mem_notatt_set = set([]) #回执不参会会员
mem_notans_set = set([]) #未回执会员
for each in mem_set:
if each in ans_attend_set:
mem_attend_set.add(each)
elif each in ans_notatt_set:
mem_notatt_set.add(each)
else:
mem_notans_set.add(each)
#3. 是否显示中间结果
def disp(ss, cap, num = True):
#ss: 名单集合
#cap: 开头描述
print(cap,'({})'.format(len(ss)))
for i in range(np.ceil(len(ss)/5).astype(int)):
pre = i * 5
nex = (i+1) * 5
#调整显示格式
dd = ''
for each in list(ss)[pre:nex]:
if len(each) == 2:
dd = dd + ' ' + each
elif len(each) == 3:
dd = dd + ' ' + each
else:
dd = dd + '' + each
print('{:3.0f} -{:3.0f} {}'.format(i*5+1,(i+1)*5,dd))
if display[0]:
disp(concil_attend_set,'\n参会理事')
disp(concil_notatt_set,'\n不参会理事')
disp(concil_notans_set,'\n未回执理事')
if display[1]:
disp(mem_attend_set,'\n参会会员')
disp(mem_notatt_set,'\n不参会会员')
disp(mem_notans_set,'\n未回执会员')
#4. 写入excel
def trans_pd(df,ss,cap,i=1):
if len(ss) == 0:
df['序号{}'.format(i)] = np.NaN
df[cap] = np.NaN
else:
df['序号{}'.format(i)] = pd.DataFrame(np.arange(len(ss))+1)
df[cap] = pd.DataFrame(list(ss))
df['_'*i]=pd.DataFrame([''])
return df
def set2list(mem_attend_set,mem_notans_set):
if len(mem_attend_set) > len(mem_notans_set):
L = len(mem_attend_set)
mem_notans_list = list(mem_notans_set)
mem_notans_list.extend([''] * (L - len(mem_notans_set)))
mem_attend_list = list(mem_attend_set)
else:
L = len(mem_notans_set)
mem_attend_list = list(mem_attend_set)
mem_attend_list.extend([''] * (L - len(mem_attend_set)))
mem_notans_list = list(mem_notans_set)
return mem_attend_list,mem_notans_list
mem_attend_list, mem_notans_list = set2list(mem_attend_set, mem_notans_set)
df1 = pd.DataFrame(mem_attend_list,columns = ['参会会员'])
df1['']=pd.DataFrame([''])
df1 = trans_pd(df1,mem_notatt_set,'不参会会员')
df1 = trans_pd(df1,mem_notans_set,'未回执会员',2)
df1.index = df1.index + 1
concil_attend_list, concil_notans_list = set2list(concil_attend_set, concil_notans_set)
df2 = pd.DataFrame(concil_attend_list,columns = ['参会理事'])
df2['']=pd.DataFrame([''])
df2 = trans_pd(df2,concil_notatt_set,'不参会理事')
df2 = trans_pd(df2,concil_notans_list,'未回执理事',2)
df2.index = df2.index + 1
writer = pd.ExcelWriter(savefile_sheet[0])
df2.to_excel(writer, sheet_name=savefile_sheet[1])
df1.to_excel(writer, sheet_name=savefile_sheet[2])
writer.save()
print('\n\n写入excel成功~~')
if __name__ == '__main__':
main(loadfile_sheet,columns,savefile_sheet,display)
version_5
version_5对修复set中出现nan的方法进行了改进和简化; 而且将程序模块化，更具可读性; 可以自由控制写入多少张表了。
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop')
print('work_directory: ', os.getcwd())
loadfile_sheet = ['理事与会员名单.xlsx','理事与会员名单']
common_columns = ['回执参加','回执不参加']
concerned_columns = ['理事','会员']
disp_columns = ['参会','不参会','未回执']
savefile_sheet = ['理事和会员回执统计.xlsx','理事回执统计','会员回执统计']
def disp(ss, cap, num = True):
#ss: 名单集合
#cap: 开头描述
print(cap,'({})'.format(len(ss)))
for i in range(np.ceil(len(ss)/5).astype(int)):
pre = i * 5
nex = (i+1) * 5
#调整显示格式
dd = ''
for each in list(ss)[pre:nex]:
if len(each) == 2:
dd = dd + ' ' + each
elif len(each) == 3:
dd = dd + ' ' + each
else:
dd = dd + '' + each
print('{:3.0f} -{:3.0f} {}'.format(i*5+1,(i+1)*5,dd))
def trans_pd(df,ss,cap,i=1):
df['_'*i]=pd.DataFrame([''])
if len(ss) == 0:
df['序号{}'.format(i)] = np.NaN
df[cap] = np.NaN
else:
df['序号{}'.format(i)] = pd.DataFrame(np.arange(len(ss))+1)
df[cap] = pd.DataFrame(list(ss))
return df
def set2list(ss1,ss2):
if len(ss1) > len(ss2):
L = len(ss1)
ss2_list = list(ss2)
ss2_list.extend([''] * (L - len(ss2)))
ss1_list = list(ss1)
else:
L = len(ss2)
ss1_list = list(ss1)
ss1_list.extend([''] * (L - len(ss1)))
ss2_list = list(ss2)
return ss1_list,ss2_list
def get_df(loadfile_sheet,common_columns,concerned_column,disp_columns, display = True):
#1. 载入excel
data = pd.read_excel(loadfile_sheet[0],loadfile_sheet[1])
common_set1 = set(data[common_columns[0]])
common_set1.discard(np.NaN)
common_set2 = set(data[common_columns[1]])
common_set2.discard(np.NaN)
concerned_set = set(data[concerned_column])
concerned_set.discard(np.NaN)
#2. 统计
concerned_in_set_1 = set([])
concerned_in_set_2 = set([])
concerned_in_no_set = set([])
for each in concerned_set:
if each in common_set1:
concerned_in_set_1.add(each)
elif each in common_set2:
concerned_in_set_2.add(each)
else:
concerned_in_no_set.add(each)
#3. 显示
if display:
disp(concerned_in_set_1,'\n'+disp_columns[0]+concerned_column)
disp(concerned_in_set_2,'\n'+disp_columns[1]+concerned_column)
disp(concerned_in_no_set,'\n'+disp_columns[2]+concerned_column)
#4. 返回DataFrame
concerned_in_set_1_list, concerned_in_set_2_list = set2list(concerned_in_set_1, concerned_in_no_set)
df = pd.DataFrame(concerned_in_set_1_list,columns = [disp_columns[0]])
df = trans_pd(df,concerned_in_set_2,disp_columns[1])
df = trans_pd(df,concerned_in_no_set,disp_columns[2],2)
df.index = df.index + 1
return df
def save2excel(df, concerned_column, savefile_sheet):
L = len(savefile_sheet) - 1
idx = 0
for i in np.arange(L)+1:
if concerned_column in savefile_sheet[i]:
idx = i
break
if idx != 0:
names = locals()
for i in np.arange(L)+1:
if i != idx:
names['df%s' % i] = pd.read_excel(savefile_sheet[0], sheet_name=savefile_sheet[i])
writer = pd.ExcelWriter(savefile_sheet[0])
for i in np.arange(L)+1:
if i != idx:
names['df%s' % i].to_excel(writer, sheet_name=savefile_sheet[i])
else:
df.to_excel(writer, sheet_name=savefile_sheet[i])
writer.save()
else:
names = locals()
for i in np.arange(L)+1:
names['df%s' % i] = pd.read_excel(savefile_sheet[0], sheet_name=savefile_sheet[i])
writer = pd.ExcelWriter(savefile_sheet[0])
for i in np.arange(L)+1:
names['df%s' % i].to_excel(writer, sheet_name=savefile_sheet[i])
df.to_excel(writer, sheet_name=concerned_column)
writer.save()
print('writing success')
if __name__ == '__main__':
for concerned_column in concerned_columns:
df = get_df(loadfile_sheet,common_columns,
concerned_column,disp_columns, display = True)
save2excel(df, concerned_column, savefile_sheet)
version_final
相比较version_5，修复了一个bug，之前需要先验知识，现在更通用一点（prep函数取代了set2list函数）。
import os
import numpy as np
import pandas as pd
os.chdir('C:\\Users\\dell\\Desktop')
print('work_directory: ', os.getcwd())
loadfile_sheet = ['理事与会员名单.xlsx','理事与会员名单']
common_columns = ['回执参加','回执不参加']
concerned_columns = ['理事','会员']
disp_columns = ['参会','不参会','未回执']
savefile_sheet = ['理事和会员回执统计.xlsx','理事回执统计','会员回执统计']
def disp(ss, cap, num = True):
#功能：显示名单
#ss : 名单集合
#cap ：开头描述
print(cap,'({})'.format(len(ss)))
for i in range(np.ceil(len(ss)/5).astype(int)):
pre = i * 5
nex = (i+1) * 5
#调整显示格式
dd = ''
for each in list(ss)[pre:nex]:
if len(each) == 2:
dd = dd + ' ' + each
elif len(each) == 3:
dd = dd + ' ' + each
else:
dd = dd + '' + each
print('{:3.0f} -{:3.0f} {}'.format(i*5+1,(i+1)*5,dd))
def trans_pd(df,ll,cap,i=1):
#功能：生成三列--空列、序号列、数据列
#df : DataFrame结构
#ll : 列表
#cap : 显示的列名
#i : 控制空列的名字
df['_'*i]=pd.DataFrame([''])
if len(set(ll)) == 1:
df['序号{}'.format(i)] = np.NaN
df[cap] = np.NaN
else:
df['序号{}'.format(i)] = pd.DataFrame(np.arange(len(set(ll))-1)+1)
df[cap] = pd.DataFrame(ll)
return df
def prep(ss, N):
#功能：预处理，生成列表，并补齐到长度N
#ss : 集体
#N ：长度
ll = list(ss)
L = len(ll)
ll.extend([np.NaN] * (N-L))
return ll
def get_df(loadfile_sheet,common_columns,concerned_column,disp_columns, display = True):
#1. 载入excel
data = pd.read_excel(loadfile_sheet[0],loadfile_sheet[1])
common_set1 = set(data[common_columns[0]])
common_set2 = set(data[common_columns[1]])
concerned_set = set(data[concerned_column])
common_set1.discard(np.NaN)
common_set2.discard(np.NaN)
concerned_set.discard(np.NaN)
#2. 统计
concerned_in_set_1 = set([])
concerned_in_set_2 = set([])
concerned_in_no_set = set([])
for each in concerned_set:
if each in common_set1:
concerned_in_set_1.add(each)
elif each in common_set2:
concerned_in_set_2.add(each)
else:
concerned_in_no_set.add(each)
#3. 显示
if display:
disp(concerned_in_set_1,'\n'+disp_columns[0]+concerned_column)
disp(concerned_in_set_2,'\n'+disp_columns[1]+concerned_column)
disp(concerned_in_no_set,'\n'+disp_columns[2]+concerned_column)
#4. 返回DataFrame
N = np.max([len(concerned_in_set_1),len(concerned_in_set_2),len(concerned_in_no_set)])
concerned_in_set_1_list = prep(concerned_in_set_1,N)
concerned_in_set_2_list = prep(concerned_in_set_2,N)
concerned_in_no_list = prep(concerned_in_no_set,N)
df = pd.DataFrame(concerned_in_set_1_list,columns = [disp_columns[0]])
df = trans_pd(df,concerned_in_set_2_list,disp_columns[1])
df = trans_pd(df,concerned_in_no_list,disp_columns[2],2)
df.index = df.index + 1
return df
def save2excel(df, concerned_column, savefile_sheet):
L = len(savefile_sheet) - 1
idx = 0
for i in np.arange(L)+1:
if concerned_column in savefile_sheet[i]:
idx = i
break
if idx != 0: #如果有对应sheet
names = locals()
for i in np.arange(L)+1:
if i != idx:
names['df%s' % i] = pd.read_excel(savefile_sheet[0], sheet_name=savefile_sheet[i])
writer = pd.ExcelWriter(savefile_sheet[0])
for i in np.arange(L)+1:
if i != idx:
names['df%s' % i].to_excel(writer, sheet_name=savefile_sheet[i])
else:
df.to_excel(writer, sheet_name=savefile_sheet[i])
writer.save()
else: #如果没有对应sheet，创建一个新sheet
names = locals()
for i in np.arange(L)+1:
names['df%s' % i] = pd.read_excel(savefile_sheet[0], sheet_name=savefile_sheet[i])
writer = pd.ExcelWriter(savefile_sheet[0])
for i in np.arange(L)+1:
names['df%s' % i].to_excel(writer, sheet_name=savefile_sheet[i])
df.to_excel(writer, sheet_name=concerned_column)
writer.save()
print('writing success')
if __name__ == '__main__':
for concerned_column in concerned_columns:
df = get_df(loadfile_sheet,common_columns,
concerned_column,disp_columns, display = True)
save2excel(df, concerned_column, savefile_sheet)
2512
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
