更改输入后的第一个函数如何停止？

Question

我试图在同一个脚本中生成两个函数—但是当我将它们导入主脚本并首先运行第一个finalGrade(grades)时，grades发生了变化，增加了一列'Final Grade'。如何防止这种情况

在第一个函数中，我使用另一个函数返回每个学生的最终成绩数据集，如下所示：

# Made Andreas Døssing Mortensen s184507 & Mads Westergaard s180799

import numpy as np
import pandas as pd
from roundGrade import roundGrade


def computeFinalGrades(grades):
    #making an array from dataFrame
    arr=grades.values
    #sorting and deleting the first two columns
    gradesarr=np.sort(np.delete(arr,(0,1),axis=1))
    #setting up a list for the final grades
    gradesInList=np.zeros(len(gradesarr))
    for i in range(len(gradesarr)):
        #if -3 is in a row, the final grade should be = -3
        if -3 in gradesarr[i]:

            gradesInList[i]= -3
        #if there is 2 or more grades in a row, execute the procedure below
        elif len(gradesarr[i])>=2:
            #Delete the smallest grade
            meangrade=np.delete(gradesarr,0,axis=1)
            #taking the mean of each row
            finalgrade = np.mean(meangrade[i])
            #making a list with all the final grades
            gradesInList[i] = roundGrade(finalgrade)

        # if there is only one value in the row, return that as the final grade 
        elif len(grades[i])==1: 
            gradesInList[i] = gradesarr
        #Setting up the dataFrame again
        df = pd.DataFrame(grades)
        #Adding final grades to DataFrame
        df['Final Grade']=gradesInList
        #Show all columns
        pd.set_option('display.max_columns', None)
        gradesFinal=df

    return gradesFinal

我的绘图函数如下所示：

#Importing add-ins
import matplotlib.pyplot as plt
import numpy as np
#import pandas as pd

#importing function
from computeFinalGrades import computeFinalGrades



def finalGrade(grades):


    #Counting number of occurencies of each grade and set labels for x,y
    computeFinalGrades(grades)['Final Grade'].value_counts().sort_index().plot("bar",title="Final grades").set(xlabel='Grades',ylabel='Count')


    #show plot
    plt.show()

    return



def assignmentGrades(grades):

    #create an array with all grades
    array = grades.values
    #Sorting and deleting "Name" and "StudentID" from the array
    gradesarray = np.sort(np.delete(array,(0,1),axis=1))
    #making a for loop, to iterate trough array
    for i in range(len(gradesarray)):
        #Defining the length of the array to define max x-values
        num_ass = len(gradesarray[0])
        #Setting the x-axis values to correspond with number of assignments + a jiggler in the interval [-.1,.1] on both axis
        x = np.arange(1,num_ass+1) + np.random.uniform(-0.1,0.1)

        y = gradesarray[i,:] + np.random.uniform(-0.1,0.1)
        #Plotting the x,y "o" for creating scatterplot
        plt.plot(x, y,"o")

    #Drawing mean of grades as line
    meangrade = np.mean(gradesarray,0)
    #Plotting the mean grade as a line
    plt.plot(x,meangrade)
    #Set labels for x,y
    plt.xlabel('Assignments')
    plt.ylabel('Grades')
    #show plot
    plt.show()

    return

希望你们能帮个忙，我好像想不通

Answer 1

我想告诉你我是如何找到解决你问题的方法的：

• 您声明了行为：grades在调用assignmentGrades之后发生更改，然后有一个额外的列Final Grade
• 因为您没有直接更改该方法中的grades，所以您可能会将其引用传递给另一个变量，然后更改它。 你已经做了两次了，
  • 用arr = grades.values和
  • 用df = pd.DataFrame(grades)一次
  • 然后将Final Grade列添加到df
• 我开始怀疑并发现grades的类型可能是一个pd.DataFrame，并且构造函数-当给定对另一个pd.DataFrame的引用时-不会复制。所以最终df与grades在同一个对象上运行。当你改变一个，你就改变了另一个

为了证明这一点，我编写了以下代码

a = pd.DataFrame({ "first": [1, 2], "second": [3, 4]})
b = pd.DataFrame(a)
# b = pd.DataFrame(a.copy()) # Fix

b['new'] = [5, 6]
print(a)
print(b)

它确实打印了更改的a和b：

   first  second  new
0      1       3    5
1      2       4    6
   first  second  new
0      1       3    5
1      2       4    6

要修复代码，请在创建df时使用grades（grades.copy()）的副本

本可以帮助您解决这个问题的是一个调试器，它允许您逐步检查代码并查看变量的当前值。这样，您就可以看到grades何时发生变化，然后得出结论

---

我还建议将df重命名为gradesFinal，因为gradesFinal = df不起任何作用