bugfix-for-issue#851

[kashish2210]

Bug Fix for Issue #851: Simulator.simple_ir: width should always be > self.dt

Description of the Bug

The simple_ir function constructs an impulse response by appending a series of ones after a certain delay. However, when the width parameter is smaller than self.dt, the calculation:

h_ones = np.ones(int(width / self.dt)) * intensity

results in an empty array (h_ones = []). This leads to an incorrect output where the expected impulse response does not contain the required 1 value.

Steps to Replicate the Bug

1. Create an instance of the class containing simple_ir.
2. Call the function with width smaller than self.dt, for example:

   width = 0.1 * self.dt
   impulse_response = sim.simple_ir(start=3, width=width, intensity=1)
   print(impulse_response)
   

3. Observe the output: Actual Output: [0, 0, 0] Expected Output: [0, 0, 0, 1]

Proposed Fix

To ensure that at least one bin is always created, we modify the h_ones calculation to:

num_bins = max(int(width / self.dt), 1)
h_ones = np.ones(num_bins, dtype=int) * intensity

How This Fix Works

• Ensures at least one bin is created in the impulse response when width < self.dt.
• Prevents an empty h_ones array, which was the root cause of the issue.
• Ensures integer output (dtype=int) to match expected results [0, 0, 0, 1] instead of [0. 0. 0. 1.].

Testing the Fix

Run the following test case:

width = 0.1 * sim.dt
impulse_response = sim.simple_ir(start=3, width=width, intensity=1)
print(impulse_response)  # Expected output: [0, 0, 0, 1]

Before Fix: [0, 0, 0] (Buggy output)

After Fix: [0, 0, 0, 1] (Correct output)

Pull Request Guidelines Followed

• [x] Followed coding standards (PEP8, NumPy conventions).
• [x] Tested the fix with a valid test case.

Preview Output: