py

Created on Mon Mar 16 11:23:05 2015@author: admin"""# 分批次运行，先得到books， 再运行，利用print 来debugimport pandas as pdimport numpy as np import matplotlib as plimport oscount=0#ori_path='C:\SecondHalf_zhu\Tasks\冯静\样本数据'ori_path='C:\Python27\data\sample'def get_xls_books(ori_path): # get target filename and dir of a path     temp = list(os.walk(ori_path)) #tmp is list ,tmp[0]该目录文件路径和文件名， tmp[1]:子目录文件路径，文件名称     book_path=[]    book_name=[]    root=temp[0][0]    for i in temp[0][2]:      if os.path.splitext(i)[1] == '.xlsx': #splite filename           obj_path=os.path.join(root,i)          book_path.append(obj_path)          book_name.append(os.path.splitext(i)[0])    return zip(book_path, book_name) def get_xls_sheets(obj_path): #get all sheets of a object path    xls=pd.ExcelFile(obj_path)      sheet_names=xls.sheet_names    sheets=[xls.parse(i)  for i in range(0,len(sheet_names)) ]   # use parse can get dataframe saved into a list     return zip(sheet_names,sheets)books=get_xls_books(ori_path)  # book paths list#xls=pd.ExcelFile(books[0])#print xls.parse(0).columns for book in books:  # this loop get a book     bookname=book[1]    bookpath=book[0]    sheets=get_xls_sheets(bookpath)  # get all sheets and names from book path     for i in range(len(sheets)):  #operate each sheet, is range(len) not len         sheetname=sheets[i][0]        name=bookname+'-'+sheetname+'.'+'csv'         pd_sheet=sheets[i][1] # dataFrame        cols=pd_sheet.columns        pdf=pd_sheet[[cols[17],cols[18],cols[4],cols[6],cols[8],cols[10],cols[19],cols[16]]]  # use emunerate function             pdf.columns=['area','patch','wangdian','operator','custid','servid','servtype','is_identify'] # rename         pdf1=pdf[pdf['is_identify']=='是']   #filter        pdf2=pdf1.drop('is_identify',1)             pdf2.to_csv(os.path.join(ori_path,name), header=None, index=None)