https://raw.githubusercontent.com/MatthewReid854/reliability/master/docs/images/logo.png

sample_size_no_failures

class reliability.Reliability_testing.sample_size_no_failures

This is used to determine the sample size required for a test in which no failures are expected, and the desired outcome is the lower bound on the reliability based on the sample size and desired confidence interval.

Parameters:
  • reliability (float) – The lower bound on product reliability. Must be between 0 and 1.
  • CI (float, optional) – The confidence interval of the result. Must be between 0.5 and 1 since a confidence less than 50% is not meaningful. Default = 0.95 for 95% CI.
  • lifetimes (int, float, optional) – If testing the product for multiple lifetimes then more failures are expected so a smaller sample size will be required to demonstrate the desired reliability (assuming no failures). Conversely, if testing for less than one full lifetime then a larger sample size will be required. Default = 1. Must be greater than 0. No more than 5 is recommended due to test feasibility.
  • weibull_shape (int, float, optional) – If the weibull shape (beta) of the failure mode is known, specify it here. Otherwise leave the default of 1 for the exponential distribution.
  • print_results (bool, optional) – If True the results will be printed to the console. Default = True.
Returns:

n (int) – The number of items required in the test. This will always be an integer (rounded up).