minte9

learningjourney

Make blobs

 
""" Make blobs
Dataset that work well with clustering techniques.
"""

import matplotlib.pyplot as plt
from sklearn.datasets import make_blobs

features, target = make_blobs(
    n_samples = 100,
    n_features = 2,
    centers = 3, # three target classes
    cluster_std = 0.5,
    shuffle = True,
    random_state = 1
)

print(features[0:3])
    # [[-1.22685609    3.25572052]
    #  [-9.57463218   -4.38310652]
    #  [-10.71976941  -4.20558148]]

print(target[:10])
    # [0 1 1 1 2 2 2 1 0 0]

plt.scatter(features[:, 0], features[:, 1], c=target)
plt.title('Make blob - Simultated dataset')
plt.show()

Make classification

 
""" Make classification
Create simulated data for classification.
"""

import matplotlib.pyplot as plt
from sklearn.datasets import make_classification

features, target = make_classification(
    n_samples = 100,
    n_features = 2,
    n_informative = 2,
    n_redundant = 0,
    n_classes = 2,
    weights = [.25, .75],
    random_state = 1
)

print(features[0:3])
    # [[ 1.30022717 -0.7856539 ]
    #  [ 1.44184425 -0.56008554]
    #  [-0.84792445 -1.36621324]]

print(target[:10])
    # [0 1 1 1 2 2 2 1 0 0]

plt.scatter(features[:, 0], features[:, 1], c=target)
plt.title('Make classification - Simultated dataset')
plt.show()

Make regression

 
""" Make regression
Dataset designed to be used with linear regression.
"""

import matplotlib.pyplot as plt
from sklearn.datasets import make_regression

# Sample dataset
features, target, coef = make_regression(
    n_samples = 100,
    n_features = 3,
    n_informative = 3,
    n_targets = 1,
    noise = 0,
    coef = True,
    random_state = 1
)

# Feature Matrix (first three rows)
print(features[:2])
    # [[ 1.29322588 -0.61736206 -0.11044703]
    #  [-2.793085    0.36633201  1.93752881]
    #  [ 0.80186103 -0.18656977  0.0465673 ]]

# Target vector (first three elements)
print(target[:3])
    # [-10.37865986  25.5124503   19.67705609]

Salary Example

 
""" Simulated dataset (experience / salary)

Simulate the data for building a regression model.
Suppose, we have a survey among the employees of a company.

As a developer, often you have no access to survey data.
You need to simulate the data for building the regression model.
"""

import matplotlib.pyplot as plt
from sklearn.datasets import make_regression
from sklearn.linear_model import LinearRegression
import numpy as np
import pandas as pd

# Sample dataset
X, y = make_regression(
    n_samples = 100, 
    n_features = 1, 
    n_informative = 1, 
    n_targets = 1,
    noise = 10, 
    coef = False, 
    random_state = 0
)
print(X[:1]) # [[-0.35955316]]
print(y[:1]) # [-19.95588561]


# Scale
X = np.interp(X, (X.min(), X.max()), (0, 20))           # years of experience
y = np.interp(y, (y.min(), y.max()), (10000, 200000))   # salary

# Dataframe
df = pd.DataFrame(data={'Experience': X.flatten(), 'Salary': y})
print(df.head(2))
    #  Experience         Salary
    #    9.096218   95224.004179
    #   14.637429  132619.663538

# Plot dataset points
plt.scatter(X, y, label='training data')
plt.title('Simultated dataset (Experience / Salary)')


# Fit a linear regression model
reg = LinearRegression().fit(X, y)

# Plot the regression line 
x_line = np.linspace(np.min(X), np.max(X), 100)
y_line = reg.intercept_ + x_line * reg.coef_[0]
plt.plot(x_line, y_line, color='red', label='prediction')
plt.text(10, 25000, r'y = %0.2f + %0.2f x' % (reg.intercept_, reg.coef_[0]))
plt.xlabel('Years of experience')
plt.ylabel('Salary')
plt.legend()
plt.show()