Skip to main content

HowTo: Visualize data distribution - Linear-Regression with TensorFlow

1. Using Jupyter Notebook (Python)​

Step 1. Sign in TWSC​


Step 2. Create an Interactive Container​

  • Please refer to Interactive Container to create an Interactive Container, and select TensorFlow for the image type (For image, please select a version with Python 2 before 19.08 (not including 19.08)).

Step 3. Connect to the Container​

  • Use Jupyter Notebook to connect the container and add a Python 2 notebook.
info

See Connection methods


Step 4. Execute Linear-Regression program​

  • Copy and paste the following code to Jupyter Notebook.
%matplotlib inline
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt

# Use numpy to generate 100 random points
x_data = np.random.rand(100).astype(np.float32)
y_data = x_data * 0.1 + 0.3

# Try to find values for W and b that compute y_data = W * x_data + b
# (We know that W should be 0.1 and b 0.3, but TensorFlow will
# figure that out for us.)

W = tf.Variable(tf.random_uniform([1], -1.0, 1.0))
b = tf.Variable(tf.zeros([1]))
y = W * x_data + b

# Minimize the mean squared errors.
loss = tf.reduce_mean(tf.square(y - y_data))
optimizer = tf.train.GradientDescentOptimizer(0.2)
train = optimizer.minimize(loss)

# Before starting, initialize the variables.  We will 'run' this first.
init = tf.global_variables_initializer()

# Launch the graph.
sess = tf.Session()
sess.run(init)

# Fit the line.
for step in range(201):
    sess.run(train)
    if step % 20 == 0:
        print(step, sess.run(W), sess.run(b))
        plt.plot(x_data, y_data, 'ro', label='Original data')
        plt.plot(x_data, sess.run(W) * x_data + sess.run(b), label='Fitted line')
        plt.legend()
        plt.show()
        
# Learns best fit is W: [0.1], b: [0.3]
  • Click Run.


Step 5. Visualize data distribution​

  • TensorFlow will slowly find the fitting weights and draw a linear regression line.
0 [-0.7029411][0.33094117]
100 [0.03479815][0.33622062]
200 [0.09321669][0.30376825]

2. Using SSH or Jupyter Notebook (Terminal) for connection​

info

💡 The following example is based on the official TensorFlow tutorial.


Step 1. Using SSH to log in or open Jupyter Notebook (Terminal)​

info

See connection methods


Step 2. Download the TensorFlow program from GitHub​

git clone https://github.com/tensorflow/tensorflow.git


Step 3. Switch to branch Tensorflow 1.10​

cd tensorflow && git checkout r1.10


Step 4. Switch to the example/regression directory​

cd tensorflow/examples/get_started/regression


Step 5. Run the sample program using Python commands​

python linear_regression.py
  • The following message will be output during the computation:
    • Check point directory: You can use the TensorBoard tool to visualize neural networks and analyze training trends
    • The loss value after every 100 iterations, which helps to determine whether the model training has converged