Python Forum

Hello, I am new in python. Need help with neural network code.

I have data in .xlxs file. Dimensions 7027x65.
This is data file
Data file

First of all I calculate 20% of my data for training my neural network, and set output training array

import pandas as pd
import numpy as np

dataset = pd.read_excel("C:\\Users\\tigra\\Desktop\\1year.xlsx") # nuskaitymas exel failo

# Calculate 20%
dydis=dataset.shape
print('Masyvo dimension: ',dydis)
row=round((dydis[0]*(dydis[1]-1))*0.2/(dydis[0]))
Training_array = dataset.iloc[:,:row]

Training_array_dimensions=Training_array.shape
print('Masyvo dimension: ',Training_array_dimensions)

# input data
inputs = Training_array
outputs = np.array([100])

For prediction I am using other 80% of data

# create two new examples to predict                                   
example = dataset.iloc[:,row:]
example_dimension= example.shape
print('Masyvo dimension: ',example_dimension)

This all the code:

# create NeuralNetwork class
class NeuralNetwork:

    # intialize variables in class
    def __init__(self, inputs, outputs):
        self.inputs  = inputs
        self.outputs = outputs
        # initialize weights as .50 for simplicity
        self.weights = np.full([7027, 1], .50) 
        self.error_history = []
        self.epoch_list = []

    #activation function ==> S(x) = 1/1+e^(-x)
    def sigmoid(self, x, deriv=False):
        if deriv == True:
            return x * (1 - x)
        return 1 / (1 + np.exp(-x))

    # data will flow through the neural network.
    def feed_forward(self):
        self.hidden = self.sigmoid(np.dot(self.inputs, self.weights))

    # going backwards through the network to update weights
    def backpropagation(self):
        self.error  = self.outputs - self.hidden
        delta = self.error * self.sigmoid(self.hidden, deriv=True)
        self.weights += np.dot(self.inputs.T, delta)

    # train the neural net for 25,000 iterations
    def train(self, epochs=25000):
        for epoch in range(epochs):
            # flow forward and produce an output
            self.feed_forward()
            # go back though the network to make corrections based on the output
            self.backpropagation()    
            # keep track of the error history over each epoch
            self.error_history.append(np.average(np.abs(self.error)))
            self.epoch_list.append(epoch)

    # function to predict output on new and unseen input data                               
    def predict(self, new_input):
        prediction = self.sigmoid(np.dot(new_input, self.weights))
        return prediction

# create neural network   
NN = NeuralNetwork(inputs, outputs)
# train neural network
NN.train()

# print the predictions for both examples                                   
print(NN.predict(example), ' - Correct: ')

By running the code I'm get this error:


How to fix that? Thanks for your help.

Please could you post errors as text using error tags, not as an image of an error.
Thank you

LOAD_GLOBAL(dispatcher), LOAD_FAST(args), LOAD_FAST(kwargs), CALL_FUNCTION_EX{1}, STORE_FAST(relevant_args)
return implement_array_function(

Error:
Exception has occurred: ValueError
shapes (7027,13) and (7027,1) not aligned: 13 (dim 1) != 7027 (dim 0)
  File "<__array_function__ internals>", line 5, in dot
  File "C:\Users\tigra\Desktop\Untitled-1.py", line 49, in feed_forward
    self.hidden = self.sigmoid(np.dot(self.inputs, self.weights))
  File "C:\Users\tigra\Desktop\Untitled-1.py", line 61, in train
    self.feed_forward()
  File "C:\Users\tigra\Desktop\Untitled-1.py", line 80, in <module>
    NN.train()

implementation, dot, relevant_args, args, kwargs)

Don't take this wrong - are you supposed to be doing it the hard way, coding each piece when there are fully vetted functions out there to do it? I know sometimes assignments are done that way to try to teach you something special, but for example:

trainval_dataset = df.sample(frac=0.8,random_state=42)
test_dataset = df.drop(trainval_dataset.index)

does your training and testing split, in the usual 80% training and 20% test.

Current problem is with the shapes of your data. Look at the shapes being fed into feedforward.