Building a Logistic Regression Model with Python & Running the Project in Docker: A Hands-On Guide

Building a Logistic Regression Model with Python: A Hands-On Guide

Logistic regression is one of the most commonly used machine learning algorithms for classification problems. In this post, we will walk through the steps to implement a simple logistic regression model on the famous Iris dataset, visualize some data characteristics, and save the trained model for future use.

Download this Dataset

Step 1: Loading the Dataset

First, we load the Iris dataset, which is stored in a CSV file. The dataset contains 150 rows and 5 columns, where each row represents a flower sample and includes measurements like Sepal Length, Sepal Width, Petal Length, and Petal Width, along with the species of the flower.

from pandas import read_csv
filename = "Iris.csv"
data = read_csv(filename)

Step 2: Displaying Dataset Information

After loading the dataset, it's essential to understand its structure. We can check the dataset's shape and preview the first 20 rows to get a feel of the data.

print("Shape of the dataset:", data.shape)
print("First 20 rows:\n", data.head(20))

Step 3: Visualizing Data with Histograms

Before building a model, it's crucial to visualize the data. Histograms give us a quick overview of the distribution of each feature. We plot and save them silently to avoid displaying them in the notebook.

from matplotlib import pyplot
data.hist()
pyplot.savefig("histograms.png")
pyplot.close()

Step 4: Density Plot for Data Distribution

Next, we plot density plots to better understand the distribution of each feature. This can help us detect skewness or identify any patterns that might be useful during training.

data.plot(kind='density', subplots=True, layout=(3,3), sharex=False)
pyplot.savefig("density_plots.png")
pyplot.close()

Step 5: Preparing the Data for Training

Now, we convert the dataset into a NumPy array and separate the features (the measurements) from the labels (species). This will allow us to feed the data into the logistic regression model.

array = data.values
X = array[:, 1:5]  # Features: Sepal/Petal measurements
Y = array[:, 5]    # Target: Species

Step 6: Splitting the Data

We split the data into training (67%) and testing (33%) sets to evaluate our model's performance. The train_test_split function from sklearn handles this.

from sklearn.model_selection import train_test_split
test_size = 0.33
seed = 7
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=test_size, random_state=seed)

Step 7: Building and Training the Logistic Regression Model

We now create a logistic regression model and train it using the training data. We set max_iter=200 to ensure the model has enough iterations to converge.

from sklearn.linear_model import LogisticRegression
model = LogisticRegression(max_iter=200)
model.fit(X_train, Y_train)

Step 8: Evaluating Model Accuracy

Once the model is trained, we evaluate its performance on the testing set. We calculate the accuracy to see how well the model generalizes.

result = model.score(X_test, Y_test)
print("Accuracy: {:.2f}%".format(result * 100))

Step 9: Saving the Trained Model

Finally, we save the trained model using joblib so we can load it later for predictions without retraining.

import joblib
joblib.dump(model, "logistic_model.pkl")

Running the Project in Docker

Prerequisite

• Download Docker Desktop
• Run the App

Create the Project Structure

Create a folder named iris_logistic_regression/ and place the following files inside:

iris_logistic_regression/
│
├── Iris.csv
├── model.py
├── requirements.txt
└── Dockerfile

Dockerfile code

FROM python:3.9-slim

RUN pip install pandas scikit-learn matplotlib

COPY . .

CMD ["python","iris.py"]

Build the Docker Image

docker build -t iamjabastin/24mcr037-ml:latest .

Check the image

docker images

Push the image

docker push iamjabastin/24mcr037-ml

Open the Docker Desktop

Run the Docker Image

Docker Image runs successfully with the output.