英文:
FOR Loop -only giving result for 1st iteration in function
问题
FOR循环仅在第一次迭代时给出结果，然后停止，不显示任何错误，
在查找pandas.dataframe的中位数时
def construct_frame(x):
'''constructor function to get median of any data'''
if x.shape[1] >= 0:
for i in range(1, (x.shape[1])):
print(i)
x_ser = x.iloc[:, i].sort_values(ignore_index=True) # 按降序排序值
x_ser_new = pd.Series(x_ser)
# print(x_ser)
# print(len(x_ser))
# print(type(x_ser))
# print(x_ser.dtypes)
# print(x_ser.shape[0])
if len(x_ser_new) % 2 == 0: # 检查数据的长度是否为'偶数'
res = (x_ser_new[len(x_ser_new) // 2] + x_ser_new[(len(x_ser_new) // 2) - 1]) / 2 # 第n列和(n-1)列的平均值
return res
else: # 如果数据的长度为'奇数'
res = x_ser_new[len(x_ser_new) // 2] # 第n列
return res
construct_frame(df)
结果应该是10列，但仅显示第1列的中位数结果：
1
34.26
英文:
FOR loop is only giving result for first iteration then it stops and doesn't show any error also,
while finding median of pandas.dataframe
def construct_frame(x):
'''constructor function to get median of any data'''
if x.shape[1]>=0:
for i in range(1,(x.shape[1])):
print(i)
x_ser = x.iloc[:,i].sort_values(ignore_index=True) #sorting values in descending order
x_ser_new= pd.Series(x_ser)
# print(x_ser)
# print(len(x_ser))
# print(type(x_ser))
# print(x_ser.dtypes)
# print(x_ser.shape[0])
if len(x_ser_new)% 2==0: #checking length of data if 'even'
res = (x_ser_new[len(x_ser_new)//2]+ x_ser_new[(len(x_ser_new)//2)-1])/2 # nth column and (n-1)th column average
return res
else: # if length of data is 'odd'
res = x_ser_new[len(x_ser_new)//2] # nth column
return res
construct_frame(df)
result should be for 10 columns but shows only 1st column median result:
1
34.26
答案1
得分: 1
不要在循环内部返回res，否则会在第一次迭代时中断循环：
def construct_frame(x):
'''constructor function to get median of any data'''
median = {} # 此处
if x.shape[1]>=0:
for i in range(1,(x.shape[1])):
x_ser = x.iloc[:,i].sort_values(ignore_index=True)
x_ser_new = pd.Series(x_ser)
if len(x_ser_new) % 2 == 0:
res = (x_ser_new[len(x_ser_new) // 2] + x_ser_new[(len(x_ser_new) // 2) - 1]) / 2
else:
res = x_ser_new[len(x_ser_new) // 2]
median[x_ser.name] = res # 此处
return pd.Series(median, name='median') # 此处
输出：
>>> construct_frame(df)
Rank 2.0
Category C
Name: median, dtype: object
更快的方法：
>>> df.iloc[:, 1:].quantile(0.5, numeric_only=False, interpolation='lower')
Rank 2.0
Category C
Name: 0.5, dtype: object
英文:
Don't return res inside the loop otherwise it breaks the loop on the first iteration:
def construct_frame(x):
'''constructor function to get median of any data'''
median = {} # HERE
if x.shape[1]>=0:
for i in range(1,(x.shape[1])):
x_ser = x.iloc[:,i].sort_values(ignore_index=True)
x_ser_new = pd.Series(x_ser)
if len(x_ser_new)% 2==0:
res = (x_ser_new[len(x_ser_new)//2]+ x_ser_new[(len(x_ser_new)//2)-1])/2
else:
res = x_ser_new[len(x_ser_new)//2]
median[x_ser.name] = res # HERE
return pd.Series(median, name='median') # HERE
Output:
>>> construct_frame(df)
Rank 2.0
Category C
Name: median, dtype: object
A faster method:
>>> df.iloc[:, 1:].quantile(0.5, numeric_only=False, interpolation='lower')
Rank 2.0
Category C
Name: 0.5, dtype: object
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论