CODEWITHSUNDEEP
pythonpandasjupyter-notebook
Jordan P
Jordan P

Reputation: 31

ValueError: could not convert string to float: 'no'

While implementing Logistic Regression on some bank data I faced an error ValueError: could not convert string to float: 'no'. Here is the code I have tried until now.

bank_full=pd.read_csv("/home/bilal/Desktop/linkedinlearning/recommendation-system-python/bank/bank-full.csv")
bank_full.head()

X=bank_full.iloc[:,:37].values 
y=bank_full.iloc[:,:18].values
Logreg=LogisticRegression()
Logreg.fit(X,y) #ERROR HERE.

enter image description here

Upvotes: 0

Views: 12443

Answers (3)

a.hilary
a.hilary

Reputation: 134

If you're using the Bank Marketing Data Set, the target(y) values are encoded as 'yes' and 'no'. You could do something like this:

bank.loc[bank.y == "yes", 'subscribe'] = 1
bank.loc[bank.y == "no", 'subscribe'] = 0

Upvotes: 1

Gavin Wong
Gavin Wong

Reputation: 1260

This error means that there are some lines in your data which do not contain valid float data, specifically a line which contains the string no.

I would suggest using a for loop to loop over your data to check which lines are the ones causing the error, as follows:

for i in data:
    try:
        i = float(i)
        print(i)
    except:
        print("Invalid data.")

Upvotes: 1

nav610
nav610

Reputation: 791

It sounds like you have a value in your data that is supposed to be a number but is actually a string called 'no'. Perhaps give the data a quick check before you convert it from strings to floats.

Upvotes: 1

Related Questions