The article presents the results of a comparative analysis of randomness testing using statistical tests in the Information Technology Laboratory of the National Institute of Standards and Technology (STS NIST) and Move-To-Front test in software implementations of floating-point random number generators and random integer numbers of the MATLAB package and the NumPy Python library. It was found thatin MATLAB package passing the randomness of Move-To-Front test of a sequence which was made up of random integer numbers calculated on the basis of the conversion from floating-point random number generators to random integer numbers depends on the sequence length and the size of the alphabet used to convert from floating-point random number generators to random integer numbers. According to the authors, as this effect is not detected in the NumPy Python library, the result obtained in the workindicates the presence of an internal defect in the algorithm for generating floating-point random number generators MATLAB.