Machine learning : Cast string to float is not supported in sequential model using python language

Asked Jun 29 '21 at 18:43

Active Jul 01 '21 at 18:23

Viewed 106 times

I try to predict the class but it shows the error:

Cast string to float is not supported.

I used model.predict_classes that is deprecated now. So I used (model.predict(x) > 0.5).astype("int32"), but I get the same error.

I use to look at some other similar questions like this, but doesn't work for me. I am using python language and code at colab.

edited Jul 01 '21 at 18:23

asked Jun 29 '21 at 18:43

yasir imran

1

The error is clear: You have some strings in your input which is the `csv` file. Check it for strings. It may be the headers that you didn't exclude it while reading. – Kaveh Jun 29 '21 at 18:54
print `xnew.dtypes` and iadd its output to the question to see if all values are in the expected format. – Kaveh Jun 29 '21 at 18:58
Here is the xnew.dtypes result xnew.dtypes ADDRESS object OPCODE object dtype: object – yasir imran Jun 29 '21 at 19:14
Should the ADDRESS column values be integer? – Kaveh Jun 29 '21 at 19:24
integer or hexadecimal – yasir imran Jun 29 '21 at 19:58
`xnew.dtypes` print `object` the type of ADDRESS. So check your book2.csv file for values of the address column. It have null values or string values. – Kaveh Jun 29 '21 at 20:26
Thank you for helping out. I change the whole data frame to float and it starts working. – yasir imran Jun 30 '21 at 08:19

0 Answers0