This software detects patterns of non-randomness, or runs, within sequences.
Skidmarks is a Python module that offers several functions to check a sequence for randomness. The module assumes some cases to be binary sequences which contain only two distinct values, not just zeroes or ones. If users have any feedback or improvements, they are encouraged to share it.
The module's functions allow you to run tests on sequences. For instance, when using the wald_wolfowitz function, you input a sequence in quotes and get a result. The wald_wolfowitz function outputs the number of runs (in the sequence), as well as the probability of the sequence's randomness. These results are conveyed as a dictionary. Users looking for evidence to reject randomness, should expect a significance value lower than 0.05.
When passing a list to wald_wolfowitz, the returned values are the same as when using a sequence in quotes. However, the returned values are different when using the autocorrelation function. Instead of running tests on values, it will look at patterns in the sequence to determine if it's random. Again, the function returns a dictionary displaying the probability of randomness.
The serial_test, gap_test, and the auto-correlation function also offer similar outputs in dictionary form. These functions provide probabilities specific to their respective area of testing.
Gap_test is unique in that you can specify the gap value initially (default set to the first value). These test results can show the likelihood of a sequence's randomness as well.
The module's functions allow you to run tests on sequences. For instance, when using the wald_wolfowitz function, you input a sequence in quotes and get a result. The wald_wolfowitz function outputs the number of runs (in the sequence), as well as the probability of the sequence's randomness. These results are conveyed as a dictionary. Users looking for evidence to reject randomness, should expect a significance value lower than 0.05.
When passing a list to wald_wolfowitz, the returned values are the same as when using a sequence in quotes. However, the returned values are different when using the autocorrelation function. Instead of running tests on values, it will look at patterns in the sequence to determine if it's random. Again, the function returns a dictionary displaying the probability of randomness.
The serial_test, gap_test, and the auto-correlation function also offer similar outputs in dictionary form. These functions provide probabilities specific to their respective area of testing.
Gap_test is unique in that you can specify the gap value initially (default set to the first value). These test results can show the likelihood of a sequence's randomness as well.
What's New
Version 0.0.3: N/A