Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userIteration in ACL2
353
0
0.0
(
0
)
Added by
EPTCS
Publication date
2020
fields
Informatics Engineering
and research's language is
English
Authors
Matt Kaufmann
Ask ChatGPT about the research
No Arabic abstract
Iterative algorithms are traditionally expressed in ACL2 using recursion. On the other hand, Common Lisp provides a construct, loop, which -- like most programming languages -- provides direct support for iteration. We describe an ACL2 analogue loop$ of loop that supports efficient ACL2 programming and reasoning with iteration.
rate research
Read More
A perfect number is a positive integer n such that n equals the sum of all positive integer divisors of n that are less than n. That is, although n is a divisor of n, n is excluded from this sum. Thus 6 = 1 + 2 + 3 is perfect, but 12 < 1 + 2 + 3 + 4 + 6 is not perfect. An ACL2 theory of perfect numbers is developed and used to prove, in ACL2(r), this bit of mathematical folklore: Even if there are infinitely many perfect numbers the series of the reciprocals of all perfect numbers converges.
Given a field K, a quadratic extension field L is an extension of K that can be generated from K by adding a root of a quadratic polynomial with coefficients in K. This paper shows how ACL2(r) can be used to reason about chains of quadratic extension fields Q = K_0, K_1, K_2, ..., where each K_i+1 is a quadratic extension field of K_i. Moreover, we show that some specific numbers, such as the cube root of 2 and the cosine of pi/9, cannot belong to any of the K_i, simply because of the structure of quadratic extension fields. In particular, this is used to show that the cube root of 2 and cosine of pi/9 are not rational.
We formalize some basic properties of Fourier series in the logic of ACL2(r), which is a variant of ACL2 that supports reasoning about the real and complex numbers by way of non-standard analysis. More specifically, we extend a framework for formally evaluating definite integrals of real-valued, continuous functions using the Second Fundamental Theorem of Calculus. Our extended framework is also applied to functions containing free arguments. Using this framework, we are able to prove the orthogonality relationships between trigonometric functions, which are the essential properties in Fourier series analysis. The sum rule for definite integrals of indexed sums is also formalized by applying the extended framework along with the First Fundamental Theorem of Calculus and the sum rule for differentiation. The Fourier coefficient formulas of periodic functions are then formalized from the orthogonality relations and the sum rule for integration. Consequently, the uniqueness of Fourier sums is a straightforward corollary. We also present our formalization of the sum rule for definite integrals of infinite series in ACL2(r). Part of this task is to prove the Dini Uniform Convergence Theorem and the continuity of a limit function under certain conditions. A key technique in our proofs of these theorems is to apply the overspill principle from non-standard analysis.
The Cayley-Dickson Construction is a generalization of the familiar construction of the complex numbers from pairs of real numbers. The complex numbers can be viewed as two-dimensional vectors equipped with a multiplication. The construction can be used to construct, not only the two-dimensional Complex Numbers, but also the four-dimensional Quaternions and the eight-dimensional Octonions. Each of these vector spaces has a vector multiplication, v_1*v_2, that satisfies: 1. Each nonzero vector has a multiplicative inverse. 2. For the Euclidean length of a vector |v|, |v_1 * v_2| = |v_1| |v2|. Real numbers can also be viewed as (one-dimensional) vectors with the above two properties. ACL2(r) is used to explore this question: Given a vector space, equipped with a multiplication, satisfying the Euclidean length condition 2, given above. Make pairs of vectors into new vectors with a multiplication. When do the newly constructed vectors also satisfy condition 2?
As Graphics Processing Units (GPUs) have gained in capability and GPU development environments have matured, developers are increasingly turning to the GPU to off-load the main host CPU of numerically-intensive, parallelizable computations. Modern GPUs feature hundreds of cores, and offer programming niceties such as double-precision floating point, and even limited recursion. This shift from CPU to GPU, however, raises the question: how do we know that these new GPU-based algorithms are correct? In order to explore this new verification frontier, we formalized a parallelizable all-pairs shortest path (APSP) algorithm for weighted graphs, originally coded in NVIDIAs CUDA language, in ACL2. The ACL2 specification is written using a single-threaded object (stobj) and tail recursion, as the stobj/tail recursion combination yields the most straightforward translation from imperative programming languages, as well as efficient, scalable executable specifications within ACL2 itself. The ACL2 version of the APSP algorithm can process millions of vertices and edges with little to no garbage generation, and executes at one-sixth the speed of a host-based version of APSP coded in C- a very respectable result for a theorem prover. In addition to formalizing the APSP algorithm (which uses Dijkstras shortest path algorithm at its core), we have also provided capability that the original APSP code lacked, namely shortest path recovery. Path recovery is accomplished using a secondary ACL2 stobj implementing a LIFO stack, which is proven correct. To conclude the experiment, we ported the ACL2 version of the APSP kernels back to C, resulting in a less than 5% slowdown, and also performed a partial back-port to CUDA, which, surprisingly, yielded a slight performance increase.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
