I dont understand the error during data preprocessing

#1

here is my code:https://drive.google.com/open?id=17VMRjbIU0u-H0gwd2BuMjEyj21Oi_Ndg

#2

Hey @shreyanarula3,
I think you have mistakenly attached a blank .ipynb file. Request you to attach your updated code and I’ll look into it right away.

#3
#4 
def classWiseData(x,y):
    data = {}
    
    for i in range(CLASSES):
        data[i] = []
        
    for i in range(x.shape[0]):
        data[y[i]].append(x[i])
    
    for k in data.keys():
        data[k] = np.array(data[k])
        
    return data

In this function, you have initialized the data dictionary with numerical keys (range(CLASSES)). While appending data, you are accessing the corresponding string value of the class (y is a vector of strings). This generates a key error as the key you are trying to access does not exist.

Hope this helps!

#5

Then why did it work in image classification project

#6

It depends on the dataset. If the dataset is already label encoded (each class is assigned a numerical value), then your code won’t generate an error. Checkout this function: Label Encoder
Encode your labels before doing any preprocessing to avoid such errors.

Hope this resolved your issue!