修复 IndexError：列出超出范围的问题

Question

我正在处理一个 .csv 文件，我编写了这段代码来计算 year 列中每个值在 csv 数据集中出现的次数。

每当我在我的个人系统上运行代码时，我都会在第 10 行 row_year = suspension[5] 收到 IndexError: list out of range，但是当我在 dataquest 站点上运行代码时，代码运行良好。

csv 数据集有 7 列，第 5 列代表年。

import csv

file  = open("nfl_suspensions_data.csv")
nfl_suspensions = list(csv.reader(file))
nfl_suspensions = nfl_suspensions[1:]

years = {}

for suspension in nfl_suspensions:
    row_year = suspension[5]
    if row_year in years:
        years[row_year] = years[row_year] + 1
    else:
        years[row_year] = 1

print(years)

Answer 1

您的数据太短 - 您索引到它后面的列表中。如果你真的有year 作为第 5 列，你应该使用column[4] 来访问它——python 索引是基于 0 的。

使用error handling:

import csv

file  = open("nfl_suspensions_data.csv")
nfl_suspensions = list(csv.reader(file))
nfl_suspensions = nfl_suspensions[1:]

years = {}

 for line_nr, suspension in enumerate(nfl_suspensions):
    try:
        row_year = suspension[5]
    except IndexError:
        # 0 based line_nr, line_nr + 1 due to removed header line  
        print("Data corrupt: less then 6 entries. Line:", line_nr+1)
        print(suspension)

        # skip this data
        continue

    if row_year in years:
        years[row_year] = years[row_year] + 1
    else:
        years[row_year] = 1

print(years)

这遵循蟒蛇Ask forgiveness not permission哲学。

---

你也应该切换到

with open("nfl_suspensions_data.csv") as file:
    nfl_suspensions = list(csv.reader(file))[1:]

这是读取文件的首选方式。见python.org - reading and writing files（见第二个代码示例块）

您也可以利用collections.defaultdict：

years = defaultdict(int) # above 

并删除周围的 if 检查

# if row_year in years:
    years[row_year] += 1  # this should work using a defaultdict(int)
# else:
#    years[row_year] = 1

或使用collections.Counter

---

更短的代码，包括完成任务的文件生成（第 [5] 行 == 第 6 列的年份）：

import csv
from collections import Counter


# Create a demo data file with errors:    
with open("nfl_suspensions_data.csv","w") as f:
    for inter in range(1,10):
        for y in range(1980,2001,inter):
            f.write(f"na,na,na,na,na,{y},na,na\n")
        # corrupt line
        f.write(f"na,na,na,na\n")


# process and count the years:
with open("nfl_suspensions_data.csv") as file:
    nfl_suspensions = list(csv.reader(file))[1:]

as_columns = list(zip(*[l for l in nfl_suspensions if len(l) > 6]))
print(Counter(as_columns[5]))

输出：

Counter({'1980': 8, '1992': 5, '1998': 5, '1986': 4, '1988': 4, '1996': 4,
         '2000': 4, '1984': 3, '1989': 3, '1990': 3, '1994': 3, '1995': 3, 
         '1982': 2, '1983': 2, '1985': 2, '1987': 2, '1981': 1, '1991': 1, 
         '1993': 1, '1997': 1, '1999': 1})

---

您的逻辑已修复，应用于上面生成的数据：

def your_code_fixed(sus):
    years = {}
    for line_nr, suspension in enumerate(sus):
        try:
            row_year = suspension[5]
        except IndexError:
            # 0 based line_nr, line_nr + 1 due to removed header line  
            print("Data corrupt: less then 6 entries. Line:", line_nr+1)
            print(suspension)

            # skip this data
            continue
        if row_year in years:
            years[row_year] = years[row_year] + 1
        else:
            years[row_year] = 1
    print(years)    

with open("nfl_suspensions_data.csv") as file:
    nfl_suspensions = list(csv.reader(file))[1:]

your_code_fixed(nfl_suspensions)

以上数据文件的输出：

Data corrupt: less then 6 entries. Line: 21
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 33
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 41
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 48
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 54
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 59
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 63
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 67
['na', 'na', 'na', 'na']
Data corrupt: less then 6 entries. Line: 71
['na', 'na', 'na', 'na']

{'1981': 1, '1982': 2, '1983': 2, '1984': 3, '1985': 2, '1986': 4, '1987': 2,
 '1988': 4, '1989': 3, '1990': 3, '1991': 1, '1992': 5, '1993': 1, '1994': 3, 
 '1995': 3, '1996': 4, '1997': 1, '1998': 5, '1999': 1, '2000': 4, '1980': 8}