代码之家 › 专栏 › 技术社区 › Rio

如何使用python过滤CSV文件中两个日期之间的行并重定向到另一个文件?

datetime csv pandas python-3.x python

Rio · 技术社区 · 6 年前

我是Python的新手。我有一个CSV文件下面的数据作为一个例子。我想跳过特定日期范围(2018-08-01到2018-08-28)之间的行,并将输出重定向到单独的CSV文件。请注意,标题“上次使用”中有空格。

NUMBER,MAIL,COMMENT,COUNT,LAST USE,PERCENTAGE,TEXTN
343,user1@example.com,"My comment","21577",2018-08-06,80.436%,
222,user2@example.com,"My comment","31181",2018-07-20,11.858%,
103,user3@example.com,"My comment",540,2018-06-14,2.013%,
341,user4@example.com,"My comment",0,N/A,0.000%,

2 回复 | 直到 6 年前

jpp 6 年前

对于熊猫来说,这很简单:

import pandas as pd

# read file
df = pd.read_csv('file.csv')

# convert to datetime
df['LAST USE'] = pd.to_datetime(df['LAST USE'])

# calculate mask
mask = df['LAST USE'].between('2018-08-01', '2018-08-28')

# output masked dataframes
df[~mask].to_csv('out1.csv', index=False)
df[mask].to_csv('out2.csv', index=False)

也可以组合布尔数组来构造 mask

m1 = df['LAST USE'] >= (pd.to_datetime('now') - pd.DateOffset(days=30))
m2 = df['LAST USE'] <= pd.to_datetime('now')
mask = m1 & m2

vicente louvet 6 年前

dict阅读器文档: https://docs.python.org/3/library/csv.html#csv.DictReader

strtime文档: https://docs.python.org/3/library/datetime.html#strftime-strptime-behavior

对于每一行,我们将日期/时间字符串转换为实际的日期/时间对象,然后python可以将其与日期范围进行比较。如果值在范围内,我们将把整行写入一个单独的CSV文件。

import datetime, csv

#define all the fieldnames in the input CSV file (for use in creating / appending to output CSV file)
fieldnames = ['NUMBER','MAIL','COMMENT','COUNT','LAST USE','PERCENTAGE','TEXTN']

#open input CSV file as readonly
with open("input.csv", "r") as fin:
    #create a CSV dictionary reader object
    csv_dreader = csv.DictReader(fin)
    #iterate over all rows in CSV dict reader
    for row in csv_dreader:
        #check for invalid Date values
        if 'N/A' not in row['LAST USE']:
            #convert date string to a date object
            datetime_val = datetime.datetime.strptime(row['LAST USE'], '%Y-%m-%d')
            #check if date falls within requested range
            if datetime_val > datetime.datetime(2018, 8, 1) and datetime_val < datetime.datetime(2018, 8, 28):                
                #if it does, open output CSV file for appending
                with open("output.csv", "a") as fout:
                    #create a csv writer object using the fieldnames defined above
                    csv_writer = csv.DictWriter(fout, fieldnames=fieldnames)
                    #write the current row (from the input CSV) to the output CSV file
                    csv_writer.writerow(row)