PyTorch is an open-source machine learning library developed by Facebook. It is used for deep neural network and natural language processing purposes.
The function torch.logspace() returns a one-dimensional tensor of steps points logarithmically spaced with base base between 
.
The output tensor is 1-D of size steps.
Syntax: torch.logspace(start, end, steps=100, base=10, out=None)
Parameters:
start: the starting value for the set of point.
end: the ending value for the set of points
steps: number of points to sample between start and end. Default: 100.
base: base of the logarithm function. Default: 10.0
out(Tensor, optional): the output tensorReturn type: A tensor
Code #1:
Python3
# Importing the PyTorch library import torch # Applying the logspace function and # storing the resulting tensor in 't' a = torch.logspace(3, 10, 5) print("a = ", a) b = torch.logspace(start =-10, end = 10, steps = 5) print("b = ", b) |
Output:
a = tensor([1.0000e+03, 5.6234e+04, 3.1623e+06, 1.7783e+08, 1.0000e+10]) b = tensor([1.0000e-10, 1.0000e-05, 1.0000e+00, 1.0000e+05, 1.0000e+10])
Code #2: Visualization
Python3
# Importing the PyTorch library import torch # Importing the NumPy library import numpy as np # Importing the matplotlib.pyplot function import matplotlib.pyplot as plt # Applying the logspace function to get a tensor of size 15 with values from -5 to 5 using base 2 a = torch.logspace(-5, 5, 15, 2) print(a) # Plotting plt.plot(a.numpy(), np.zeros(a.numpy().shape), color = 'red', marker = "o") plt.title("torch.linspace") plt.xlabel("X") plt.ylabel("Y") plt.show() |
Output:
tensor([3.1250e-02, 5.1271e-02, 8.4119e-02, 1.3801e-01, 2.2643e-01, 3.7150e-01,
6.0951e-01, 1.0000e+00, 1.6407e+00, 2.6918e+00, 4.4164e+00, 7.2458e+00,
1.1888e+01, 1.9504e+01, 3.2000e+01])
[torch.FloatTensor of size 15]
