主页 / 问题 / 1409268
Accepted
Наталья
Asked:2022-07-13 05:53:51 +0800 CST2022-07-13 05:53:51 +0800 CST

按 csv 文件中的列对数据进行分组

  • 772

大家好！你能告诉我如何在csv中对值进行分组吗？我举个例子。csv 文件包含带有城市名称的列： Москва Калининград Курск

您需要获取按列分组的表：

Москва Калининград Курск 
  2         3        1
  3         1        1
  4         2        3
  5         0        0

作为函数执行的结果，我得到以下值：

Москва, 2
Калининград, 3
Курск, 1

我在网上没有找到合适的，但我自己也想不出来（经验和知识不够）。我请求你的帮助！资源：

     import csv 

     with open(f'tmp.csv', 'w') as file:
        writer = csv.writer(file)
        writer.writerow(('Москва','Калининград', 'Курск'))
        
        data = members
        json_str = json.dumps(data)
        resp = json.loads(json_str)

        def count(cities, raw):
            result = {}
            for city in cities:
                result[city] = raw.count(city)
            return result

        data = resp
        raw = []
        uniq_city = set()

        for item in data['items']:
            try:
                city = item['city']['title']
            except KeyError:
                continue
            raw.append(city)
            uniq_city.add(city)
        result = count(uniq_city, raw)

        for key, value in result.items():
            key_city = (f'{key}')
            value_city=(f'{value}')             
            
            with open(f'tmp.csv', 'a', encoding='utf-8', newline='') as file:
                writer = csv.writer(file,delimiter=",",lineterminator="\n")
                if key_city == 'Москва':
                    writer.writerow((value_city))
                elif key_city == 'Калининград':
                    writer.writerow((value_city))             
                elif key_city == 'Курск':
                    writer.writerow((value_city))
                else:
                    pass
python pandas

2 个回答

  1. 据我了解，它应该是这样的：

    import pandas as pd
df = pd.read_csv('tmp.csv') # читаем csv

# Преобразовываем файл для более удобной обработки
temp = [] 
# temp = [('Москва', 2), ('Москва', 3), ('Москва', 5), ('Калининград', 3), ('Калининград', 1), ('Калининград', 0), ('Курск', 1), ('Курск', 1), ('Курск', 0)]
for ele in df.columns:
  for val in df[ele]:
    row = (ele, val)
    temp.append(row)

# создаем пустой датафрейм со столбцами "city" и "value"
clean_df = pd.DataFrame(temp, columns=['city', 'value'])
# группируем получившийся датафрейм по город
result = clean_df.groupby(["city"]).sum()
result.to_csv('result.csv')
    • 1
  2. Best Answer

    解决方案：

    import datetime
import pandas as pd

filename='./results/tmp.csv'
cur_time = datetime.datetime.now().strftime('%d_%m_%Y_%H_%M')
df = pd.read_csv(filename, sep=',', encoding ='utf-8')

temp = [] 

for ele in df.columns:
for val in df[ele]:
    row = (ele, val)
    temp.append(row)

clean_df = pd.DataFrame(temp, columns=['city', 'value'])
clean_df = pd.cov = pd.read_csv(filename, sep=',', header=0)

result = clean_df.groupby(["city"]).mean()

total = result.T
total.to_csv(filename)
    • 0

相关问题