Python: data argument can't be an iterator

user_4943236 0
Viewed 13591

I'm trying to replicate the code that is provided here: https://github.com/IdoZehori/Credit-Score/blob/master/Credit%20score.ipynb

The function given below fails to run and give error. Can someone help me resolving it

def replaceOutlier(data, method = outlierVote, replace='median'):
'''replace: median (auto)
            'minUpper' which is the upper bound of the outlier detection'''
vote = outlierVote(data)
x = pd.DataFrame(zip(data, vote), columns=['annual_income', 'outlier'])
if replace == 'median':
    replace = x.debt.median()
elif replace == 'minUpper':
    replace = min([val for (val, vote) in list(zip(data, vote)) if vote == True])
    if replace < data.mean():
        return 'There are outliers lower than the sample mean'
debtNew = []
for i in range(x.shape[0]):
    if x.iloc[i][1] == True:
        debtNew.append(replace)
    else:
        debtNew.append(x.iloc[i][0])

return debtNew

Function Call:

incomeNew = replaceOutlier(df.annual_income, replace='minUpper')

Error: x = pd.DataFrame(zip(data, vote), columns=['annual_income', 'outlier']) TypeError: data argument can't be an iterator

PS: I understand this has been asked before, but I tried using the techniques however the error still remains

pandas numpy python
5 Answers

This actually works in pandas version 0.24.2 without having to use list around zip

user_37414420

write like this 

coef = DataFrame(list(zip(x.columns,np.transpose(log_model.coef_))))
user_100316700

This happens because of data type issue, you can first convert it to list and then that list to dataframe. Ex. pd.DataFrame(list(data)) should work.

user_81926530

zip(list1,list2) works in Jupyter Notebook, but I find that list(zip(list1,list2)) is required to work in Python's default compiler.

user_118061160
Related