Pandas CSV 平均和排序

Question

对于我最后的 Python 计算任务，我被要求用 Python 编写一个数据库程序，这将允许我访问三个班级数据库，每个班级数据库都包含一个参加算术测验的学生的三个分数。必须对代码进行三种排序方式；按字母顺序使用名字，作为平均值，将所有三个分数相加并除以三以找到唯一值，然后将分数从最高分到最低分排序。 因此假设以下是否为 CSV 文件之一：

name1       name2 score1 score2 score3
Atticus     Finch 9      8      10
Jem         Finch 5      7      6
Jean Louise Finch 3      2      4

如果最终用户希望它按字母顺序排序，这就是它在 Python IDLE GUI 上的样子：

Atticus     Finch 9      8      10
Jean Louise Finch 3      2      4
Jem Finch   Finch 5      7      6

如果最终用户希望将其按平均排序，则应如下所示：

Atticus     Finch 9
Jem         Finch 6
Jean Louise Finch 3

如果最终用户希望它从高到低排序，它应该是这样的：

Atticus     Finch 10     9      8
Jem         Finch 7      6      5
Jean Louise Finch 4      3      2

现在这是我的代码当前的样子：

print("Welcome to the Database sorter. The system works based on the following functions. Choose your class by inputting a letter, and choose the method of sorting the data by inputing a number afterwards. A is for Class A, B is for Class B and C is the Class C.1 is for soritng the data as an average, 2 is for sorting the data in alphabetical order and 3 is for sorting the data from highest to lowest.")

classanddatasorter =''
while classanddatasorter not in ["A1","A2","A3","B1","B2","B3","C1","C2","C3"]:
classanddatasorter = input("You have the following nine options. Input A1 to sort the results of Class A as an average. Input A2 to sort the results of Class A in alphabetical order. Input A3 to sort the results of Class A from highest to lowest. Input B1 to sort the results of Class B as an average. Input B2 to sort the results of Class B in alphabetical order. Input B3 to sort the results of Class B from highest to lowest. Input C1 to sort the results of Class C as an average. Input C2 to sort the results of Class C in alphabetical order. Input C3 to sort the results of Class C from highest to lowest. ")
if classanddatasorter == "A1":
 df = pd.read_csv('classa.csv')
 df[["score1", "score2","score3"]].mean(axis=1)

elif classanddatasorter == "A2":
 df = pd.read_csv('classa.csv')
 saved_column = df.column_name
 name = df.name
 name.sort 

elif classanddatasorter == "A3":
 df = pd.read_csv('classa.csv')
 df.sort[('score1','score2','score3'], ascending=False) 

elif classanddatasorter == "B1":
 df = pd.read_csv('classb.csv')
 df[["score1", "score2","score3"]].mean(axis=1)  

elif classanddatasorter == "B2":
 df = pd.read_csv('classb.csv')
 saved_column = df.column_name
 name = df.name

elif classanddatasorter == "B3":
 df = pd.read_csv('classb.csv')
 df.sort[('score1','score2','score3'], ascending=False)

elif classanddatasorter == "C1":
 df = pd.read_csv('classc.csv')
 df[["score1", "score2","score3"]].mean(axis=1)

elif classanddatasorter == "C2":
 bamboo = pd.read_csv('classc.csv')
 saved_column = df.column_name
 name = df.name
 name.sort 

elif classanddatasorter == "C3":
 df = pd.read_csv('classc.csv')
 df.sort[('score1','score2','score3'], ascending=False)

到目前为止，我收到以下错误：

尝试对代码进行平均排序：

 Traceback (most recent call last):
  File "C:\Users\MVMCJK\Downloads\Python code\Seperate independent draft of Task 3 (not intergated with Task 1 and 2) draft 3.py", line 70, in <module>
df[["score1", "score2","score3"]].mean(axis=1)
  File "C:\Users\MVMCJK\Anaconda3\lib\site-packages\pandas\core\frame.py", line 1791, in __getitem__
return self._getitem_array(key)
  File "C:\Users\MVMCJK\Anaconda3\lib\site-packages\pandas\core\frame.py", line 1835, in _getitem_array
indexer = self.ix._convert_to_indexer(key, axis=1)
  File "C:\Users\MVMCJK\Anaconda3\lib\site-packages\pandas\core\indexing.py", line 1112, in _convert_to_indexer
raise KeyError('%s not in index' % objarr[mask])
KeyError: "['score1' 'score2' 'score3'] not in index"

尝试按字母顺序对代码进行排序：

Traceback (most recent call last):
  File "C:\Users\MVMCJK\Downloads\Python code\Seperate independent draft of Task 3 (not intergated with Task 1 and 2) draft 3.py", line 74, in <module>
saved_column = df.column_name
  File "C:\Users\MVMCJK\Anaconda3\lib\site-packages\pandas\core\generic.py", line 2150, in __getattr__
(type(self).__name__, name))
AttributeError: 'DataFrame' object has no attribute 'column_name'

最后一部分甚至不能远程工作：由于语法无效，它默认拒绝运行，我必须消除它才能使程序正常工作，当我输入 A3 时它甚至没有给出响应。 我已经尝试用谷歌搜索 KeyError 和 AttributeError ，但我找不到任何与我的问题相关的东西，并使我能够找到进一步的修复。有人知道我的程序有什么好玩的吗？任何帮助将不胜感激。

编辑：更新后仍然无法正常工作的代码：

print("Welcome to the Database sorter. The system works based on the following functions. Choose your class by inputting a letter, and choose the method of sorting the data by inputing a number afterwards. A is for Class A, B is for Class B and C is the Class C.1 is for soritng the data as an average, 2 is for sorting the data in alphabetical order and 3 is for sorting the data from highest to lowest.")
classanddatasorter =''
while classanddatasorter not in ["A1","A2","A3","B1","B2","B3","C1","C2","C3"]:
classanddatasorter = input("You have the following nine options. Input A1 to sort the results of Class A as an average. Input A2 to sort the results of Class A in alphabetical order. Input A3 to sort the results of Class A from highest to lowest. Input B1 to sort the results of Class B as an average. Input B2 to sort the results of Class B in alphabetical order. Input B3 to sort the results of Class B from highest to lowest. Input C1 to sort the results of Class C as an average. Input C2 to sort the results of Class C in alphabetical order. Input C3 to sort the results of Class C from highest to lowest. ")
if classanddatasorter == "A1":
 df = pd.read_csv('classa.csv')
 df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)

elif classanddatasorter == "A2":
 df = pd.read_csv('classa.csv', index_col='name1')
 saved_column = df.column_name
 name = df.name
 name.sort 

elif classanddatasorter == "A3":
 df = pd.read_csv('classa.csv')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)


elif classanddatasorter == "B1":
 df = pd.read_csv('classb.csv')
 df['average'] = df[["score1", "score2","score3"]].mean(axis=1)


elif classanddatasorter == "B2":
 df = pd.read_csv('classb.csv',index_col='name1')
 saved_column = df.column_name
 name = df.name

elif classanddatasorter == "B3":
 df = pd.read_csv('classb.csv')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)

elif classanddatasorter == "C1":
 df = pd.read_csv('classc.csv')
 df['average'] = df[["score1", "score2","score3"]].mean(axis=1)

elif classanddatasorter == "C2":
 df = pd.read_csv('classc.csv',index_col='name1')
 saved_column = df.column_name
 name = df.name
 df = name.sort 

elif classanddatasorter == "C3":
 df = pd.read_csv('classc.csv')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)

编辑 2：更新了一些 bakkal 的代码示例。

print("Welcome to the Database sorter. The system works based on the following functions. Choose your class by inputting a letter, and choose the method of sorting the data by inputing a number afterwards. A is for Class A, B is for Class B and C is the Class C.1 is for soritng the data as an average, 2 is for sorting the data in alphabetical order and 3 is for sorting the data from highest to lowest.")
classanddatasorter =''
while classanddatasorter not in ["A1","A2","A3","B1","B2","B3","C1","C2","C3"]:
 classanddatasorter = input("You have the following nine options. Input A1 to sort the results of Class A as an average. Input A2 to sort the results of Class A in alphabetical order. Input A3 to sort the results of Class A from highest to lowest. Input B1 to sort the results of Class B as an average. Input B2 to sort the results of Class B in alphabetical order. Input B3 to sort the results of Class B from highest to lowest. Input C1 to sort the results of Class C as an average. Input C2 to sort the results of Class C in alphabetical order. Input C3 to sort the results of Class C from highest to lowest. ")

if classanddatasorter == "A1":
 df = pd.read_csv('classa.csv')
 print('Sorted by name1')
 df.sort('name1')
 print(df)
elif classanddatasorter == "A2":
 df = pd.read_csv('classa.csv')
 print('Sorted by average column')
 df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)
 print(df)
 print(df[['name1', 'name2', 'average']].sort('average'))
elif classanddatasorter == "A3":
 df = pd.read_csv('classa.csv')
 print('Sorted scores')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)

 for i in xrange(0, scores.shape[1]):
     column_name = 'rank{}'.format(i)
     df[column_name] = scores[:, i]

print(df[['name1', 'name2', 'rank2', 'rank1', 'rank0']])
elif classanddatasorter == "B1":
 df = pd.read_csv('classb.csv')
 print('Sorted by name1')
 df.sort('name1')
 print(df)
elif classanddatasorter == "B2":
 df = pd.read_csv('classb.csv')
 print('Sorted by average column')
 df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)
 print(df)
 print(df[['name1', 'name2', 'average']].sort('average'))
elif classanddatasorter == "B3":
 df = pd.read_csv('classb.csv')
 print('Sorted scores')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)

for i in xrange(0, scores.shape[1]):
    column_name = 'rank{}'.format(i)
    df[column_name] = scores[:, i]

print(df[['name1', 'name2', 'rank2', 'rank1', 'rank0']])
elif classanddatasorter == "C1":
 df = pd.read_csv('classc.csv')
 print('Sorted by name1')
 df.sort('name1')
 print(df)
elif classanddatasorter == "C2":
 df = pd.read_csv('classc.csv')
 print('Sorted by average column')
 df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)
 print(df)
 print(df[['name1', 'name2', 'average']].sort('average'))
elif classanddatasorter == "C3":
 df = pd.read_csv('classc.csv')
 print('Sorted scores')
 scores = df[['score1', 'score2', 'score3']].values
 scores.sort(axis=1)

 for i in xrange(0, scores.shape[1]):
     column_name = 'rank{}'.format(i)
     df[column_name] = scores[:, i]

print(df[['name1', 'name2', 'rank2', 'rank1', 'rank0']]) 

Answer 1

解析与探索

假设我们有一个这样的 CSV 文件（注意逗号后面的空格，并用逗号分隔，否则您需要使用 CSV 选项来适应您的特定格式）

scores.csv

name1,name2,score1,score2,score3
Atticus,Finch,9,8,10
Jem,Finch,5,7,6
Jean Louise,Finch,3,2,4

我们读取了 CSV 文件

df = pd.read_csv('scores.csv')

现在df 是：

         name1  name2  score1  score2  score3
0      Atticus  Finch       9       8      10
1          Jem  Finch       5       7       6
2  Jean Louise  Finch       3       2       4

而df.columns 是：

Index([u'name1', u'name2', u'score1', u'score2', u'score3'], dtype='object')

如您所见，df 具有 columns 但没有 column_name 属性，因此您的错误如下

AttributeError: 'DataFrame' 对象没有属性 'column_name'

排序

现在让我们进行字母排序

df.sort('name1')

结果是：

         name1  name2  score1  score2  score3
0      Atticus  Finch       9       8      10
2  Jean Louise  Finch       3       2       4
1          Jem  Finch       5       7       6

你想要平均值，让我们添加一列

df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)

df 现在有一个新列可供您排序！

         name1  name2  score1  score2  score3  average
0      Atticus  Finch       9       8      10        9
1          Jem  Finch       5       7       6        6
2  Jean Louise  Finch       3       2       4        3

如果您只想查看average 列

df[['name1', 'name2', 'average']].sort('average')


         name1  name2  average
0      Atticus  Finch        9
1          Jem  Finch        6
2  Jean Louise  Finch        3

你想要的最后一个分数排序有点棘手，因为数据不是整齐/标准化的，但这是一个尝试

scores = df[['score1', 'score2', 'score3']].values

scores 现在看起来像这样

array([[ 9,  8, 10],
       [ 5,  7,  6],
       [ 3,  2,  4]])

我们对scores 数组进行排序

scores.sort(axis=1)

array([[ 8,  9, 10],
       [ 5,  6,  7],
       [ 2,  3,  4]])

这些是您想要的排序分数，所以让我们将它们放入我们的df，我们必须为每个分数列执行此操作，因此我们可以使用scores.shape[1]，它是该二维数组中的列数

for i in xrange(0, scores.shape[1]):
    column_name = 'rank{}'.format(i)
    df[column_name] = scores[:, i]

现在我们的df 看起来像这样

         name1  name2  score1  score2  score3  rank0  rank1  rank2
0      Atticus  Finch       9       8      10      8      9     10
1          Jem  Finch       5       7       6      5      6      7
2  Jean Louise  Finch       3       2       4      2      3      4

并获得您想要的显示

df[['name1', 'name2', 'rank2', 'rank1', 'rank0']]


         name1  name2  rank2  rank1  rank0
0      Atticus  Finch     10      9      8
1          Jem  Finch      7      6      5
2  Jean Louise  Finch      4      3      2

整理数据

您可以阅读this PDF paper了解更多关于整理数据的信息

基本上很多操作会更容易，例如您的数据应如下所示

name, test, score
bob, 1, 10
bob, 2, 9

而不是

name, score1, score2
bob, 10, 9

Python 脚本

import pandas as pd
df = pd.read_csv('scores.csv')

print('Original Data')
print(df)

print('Sorted by name1')
df.sort('name1')
print(df)

print('Sorted by average column')
df['average'] = df[['score1', 'score2', 'score3']].mean(axis=1)
print(df)
print(df[['name1', 'name2', 'average']].sort('average'))

print('Sorted scores')
scores = df[['score1', 'score2', 'score3']].values
scores.sort(axis=1)

for i in xrange(0, scores.shape[1]):
    column_name = 'rank{}'.format(i)
    df[column_name] = scores[:, i]

print(df[['name1', 'name2', 'rank2', 'rank1', 'rank0']])

除了print()，您还可以将生成的数据帧保存到另一个.csv，例如.to_csv('score_sorted_avg.csv')