Min Max Normalization Python and Matlab – Data Mining

Min Max Normalization Python and Matlab – Data Mining

Min Max Normalization in Python and Matlab is today topic of discussion in this tutorial. Min-Max

normalization is very helpful in data mining, mathematics, and statistics. Hopefully, you will get benefit from this.

Data Before and After Normalization

Let’s see in the figure, the data before and after min-max normalization.

Min Max Normalization Python Source Code

Lets see the source code of Min Max Normalization in Python.

def __normalize(self , data ) :
# Save the Real shape of the Given Data
shape = data.shape
# Smoothing the  Given Data Valuesto 1 dimension
data = np.reshape( data , (-1 , ) )
# Find MinValue and MaxValue
MaxValue = np.max( data )
MinValue = np.min( data )
# Normalized values are store in a newly created array
normalized_values = list()
# Iterate through every value in data
for AttributeValue in the given data:
# Normalize
AttributeValue_normalized = (AttributeValue – MinValue ) / ( MaxValue – MinValue )
# Append it in the array
normalized_values.append( AttributeValue_normalized )
# Convert to numpy array
n_array = np.array( normalized_values )
# Reshape the array to its Real shape and return it.
return np.reshape( n_array , shape )

Explanation of the code

# Save the Real shape of the Given Data
shape = data.shape
# Smoothing the  Given Data Values to 1 dimension

data = np.reshape( data , (-1 , ) )

Some further steps:

1. We need to Save the Real shape of the data.
2. We need to smooth the given data.
3. The data is reshaped to a single-dimension.

# Find MinValue and MaxValue
MaxValue = np.max( data )
MinValue = np.min( data )

1. Then, we find the MinValue and MaxValue of the data.

normalized_values = list()
# Iterate through every value in data
for AttributeValue in the given data:
# Normalize
AttributeValue_normalized = (AttributeValue – MinValue ) / ( MaxValue – MinValue )
# Append it in the array
normalized_values.append( AttributeValue_normalized )
5. After normalization, we can Save it in the normalized_values list.

# Convert to numpy array
n_array = np.array( normalized_values )
# Reshape the array to its Real shape and return it.
return np.reshape( n_array , shape )