Discovering Math APIs by Mining Unit Tests

In today's API-rich world, programmer productivity depends heavily on the programmer's ability to discover the required APIs. In this paper, we present a technique and tool, called MATHFINDER, to discover APIs for mathematical computations by mining unit tests of API methods. Given a math expression, MATHFINDER synthesizes pseudo-code to compute the expression by mapping its subexpressions to API method calls. For each subexpression, MATHFINDER searches for a method such that there is a mapping between method inputs and variables of the subexpression. The subexpression, when evaluated on the test inputs of the method under this mapping, should produce results that match the method output on a large number of tests. We implemented MATHFINDER as an Eclipse plugin for discovery of third-party Java APIs and performed a user study to evaluate its effectiveness. In the study, the use of MATHFINDER resulted in a 2x improvement in programmer productivity. In 96% of the subexpressions queried for in the study, MATHFINDER retrieved the desired API methods as the top-most result. The top-most pseudo-code snippet to implement the entire expression was correct in 93% of the cases. Since the number of methods and unit tests to mine could be large in practice, we also implement MATHFINDER in a MapReduce framework and evaluate its scalability and response time.

Mining Unit Tests for Discovery and Migration of Math APIs

Today's programming languages are supported by powerful third-party APIs. For a given application domain, it is common to have many competing APIs that provide similar functionality. Programmer productivity therefore depends heavily on the programmer's ability to discover suitable APIs both during an initial coding phase, as well as during software maintenance. The aim of this work is to support the discovery and migration of math APIs. Math APIs are at the heart of many application domains ranging from machine learning to scientific computations. Our approach, called MATHFINDER, combines executable specifications of mathematical computations with unit tests (operational specifications) of API methods. Given a math expression, MATHFINDER synthesizes pseudo-code comprised of API methods to compute the expression by mining unit tests of the API methods. We present a sequential version of our unit test mining algorithm and also design a more scalable data-parallel version. We perform extensive evaluation of MATHFINDER (1) for API discovery, where math algorithms are to be implemented from scratch and (2) for API migration, where client programs utilizing a math API are to be migrated to another API. We evaluated the precision and recall of MATHFINDER on a diverse collection of math expressions, culled from algorithms used in a wide range of application areas such as control systems and structural dynamics. In a user study to evaluate the productivity gains obtained by using MATHFINDER for API discovery, the programmers who used MATHFINDER finished their programming tasks twice as fast as their counterparts who used the usual techniques like web and code search, IDE code completion, and manual inspection of library documentation. For the problem of API migration, as a case study, we used MATHFINDER to migrate Weka, a popular machine learning library. Overall, our evaluation shows that MATHFINDER is easy to use, provides highly precise results across several math APIs and application domains even with a small number of unit tests per method, and scales to large collections of unit tests.

On the Parameterized Complexity of Finding Separators with Non-Hereditary Properties

We study the problem of finding small s-t separators that induce graphs having certain properties. It is known that finding a minimum clique s-t separator is polynomial-time solvable (Tarjan in Discrete Math. 55:221-232, 1985), while for example the problems of finding a minimum s-t separator that induces a connected graph or forms an independent set are fixed-parameter tractable when parameterized by the size of the separator (Marx et al. in ACM Trans. Algorithms 9(4): 30, 2013). Motivated by these results, we study properties that generalize cliques, independent sets, and connected graphs, and determine the complexity of finding separators satisfying these properties. We investigate these problems also on bounded-degree graphs. Our results are as follows: Finding a minimum c-connected s-t separator is FPT for c=2 and W1]-hard for any ca parts per thousand yen3. Finding a minimum s-t separator with diameter at most d is W1]-hard for any da parts per thousand yen2. Finding a minimum r-regular s-t separator is W1]-hard for any ra parts per thousand yen1. For any decidable graph property, finding a minimum s-t separator with this property is FPT parameterized jointly by the size of the separator and the maximum degree. Finding a connected s-t separator of minimum size does not have a polynomial kernel, even when restricted to graphs of maximum degree at most 3, unless .

On the relationship between finger width, velocity, and fluxes in thermohaline convection

Double-diffusive finger convection occurs in many natural processes.The theories for double-diffusive phenomena that exist at present consider systems with linear stratification in temperature and salinity. The double-diffusive systems with step change in salinity and temperature are, however, not amenable to simple stability analysis. Hence factors that control the width of the finger, velocity, and fluxes in systems that have step change in temperature and salinity have not been understood so far. In this paper we provide new physical insight regarding factors that influence finger convection in two-layer double-diffusive system through two-dimensional numerical simulations. Simulations have been carried out for density stability ratios (R-rho) from 1.5 to 10. For each density stability ratio, the thermal Rayleigh number (Ra-T) has been systematically varied from 7x10(3) to 7x10(8). Results from these simulations show how finger width, velocity, and flux ratios in finger convection are interrelated and the influence of governing parameters such as density stability ratio and the thermal Rayleigh number. The width of the incipient fingers at the time of onset of instability has been shown to vary as Ra-T-1/3. Velocity in the finger varies as Ra(T)1/3/R-rho. Results from simulation agree with the scale analysis presented in the paper. Our results demonstrate that wide fingers have lower velocities and flux ratios compared to those in narrow fingers. This result contradicts present notions about the relation between finger width and flux ratio. A counterflow heat-exchanger analogy is used in understanding the dependence of flux ratio on finger width and velocity.

Axially homogeneous, zero mean flow buoyancy-driven turbulence in a vertical pipe

We report an experimental study of a new type of turbulent flow that is driven purely by buoyancy. The flow is due to an unstable density difference, created using brine and water, across the ends of a long (length/diameter = 9) vertical pipe. The Schmidt number Sc is 670, and the Rayleigh number (Ra) based on the density gradient and diameter is about 10(8). Under these conditions the convection is turbulent, and the time-averaged velocity at any point is `zero'. The Reynolds number based on the Taylor microscale, Re-lambda, is about 65. The pipe is long enough for there to be an axially homogeneous region, with a linear density gradient, about 6-7 diameters long in the midlength of the pipe. In the absence of a mean flow and, therefore, mean shear, turbulence is sustained just by buoyancy. The flow can be thus considered to be an axially homogeneous turbulent natural convection driven by a constant (unstable) density gradient. We characterize the flow using flow visualization and particle image velocimetry (PIV). Measurements show that the mean velocities and the Reynolds shear stresses are zero across the cross-section; the root mean squared (r.m.s.) of the vertical velocity is larger than those of the lateral velocities (by about one and half times at the pipe axis). We identify some features of the turbulent flow using velocity correlation maps and the probability density functions of velocities and velocity differences. The flow away from the wall, affected mainly by buoyancy, consists of vertically moving fluid masses continually colliding and interacting, while the flow near the wall appears similar to that in wall-bound shear-free turbulence. The turbulence is anisotropic, with the anisotropy increasing to large values as the wall is approached. A mixing length model with the diameter of the pipe as the length scale predicts well the scalings for velocity fluctuations and the flux. This model implies that the Nusselt number would scale as (RaSc1/2)-Sc-1/2, and the Reynolds number would scale as (RaSc-1/2)-Sc-1/2. The velocity and the flux measurements appear to be consistent with the Ra-1/2 scaling, although it must be pointed out that the Rayleigh number range was less than 10. The Schmidt number was not varied to check the Sc scaling. The fluxes and the Reynolds numbers obtained in the present configuration are Much higher compared to what would be obtained in Rayleigh-Benard (R-B) convection for similar density differences.

Energy-momentum conserving algorithm for nonlinear transient analysis within the framework of hybrid elements

This work deals with the formulation and implementation of an energy-momentum conserving algorithm for conducting the nonlinear transient analysis of structures, within the framework of stress-based hybrid elements. Hybrid elements, which are based on a two-field variational formulation, are much less susceptible to locking than conventional displacement-based elements within the static framework. We show that this advantage carries over to the transient case, so that not only are the solutions obtained more accurate, but they are obtained in fewer iterations. We demonstrate the efficacy of the algorithm on a wide range of problems such as ones involving dynamic buckling, complicated three-dimensional motions, et cetera.

Magnetic-properties of quasi-2-dimensional la1-xsr1+xmno4 and the evolution of itinerant electron ferromagnetism in the sro.(la1-xsrxmno3)n system

Quasi-two-dimensional oxides of the La,+,Sr,+,Mn04 system, possessing the KZNiF4 structure, show no evidence for ferromagnetic ordering in contrast to the corresponding three-dimensional La,+.Sr,MnO~ perovskites. Instead, there is an increasing tendency toward antiferromagnetic ordering with mcreasmg x m La,+,Sr,,, MnOp. Furthermore, these oxides are relatively high-resistivity materials over the entire compositional range. Substitution of Ba for Sr in La&r,.5Mn04 decreases the ferromagnetic interaction. Increasing the number of perovskite layers in SrO (La,-,Sr,MnO& causes an increase in electrical conductivity as well as ferromagnetic interaction. The oxide becomes a highly conducting ferromagnet when n 2 2.

Relation between high-tc superconductivity and structure - a study of la3-xba3+xcu6o14+delta and yba2cu3o7-delta systems

Superconductivity is found in tetragonal La3−x Ba3+x Cu6O14+δ and La, Ba)6−x Sr x Cu6O14+δ even though they do not possess Cu-O chains or the K2NiF4 structure. Resistivity measurements confirm the occurrence of a transformation from chain-superconductivity to sheet-superconductivity in YBa2Cu3O7−δ as δ is varied in the range 0.0–0.5.

High-temperature superconductivity in La- and Lu-substituted YBa2,Cu3,O7, and related oxides

The perovskites, Y0.75La0.25Ba2Cu3O7 and Y0.75Lu0.25Ba2Cu3O7, show high-Tc superconductivity (with zero resistance at or above 80 K), just as the parent compound YBa2Cu3O7. The Lu-substituted oxide, with the smallest unit-cell parameters, shows the highest Tc besides exhibiting a 100% Meissner effect. Hc1, in these oxides is around 25 mT, but the Hc2, is large. The thermopower of YBa2Cu3O7 shows a sharp transition to zero at the superconducting transition, reinforcing the bulk nature of the superconductivity. Preliminary studies show that ErBa2Cu3O7 and Er0.5Y0.5Ba2Cu3O7 are both high-temperature superconductors with zero resistance in the 82-90 K range.

High-Tc superconductivity in the Er---Ba---Cu---O and related systems

ErBa2Cu3O7 and Er0.5Y0.5Ba2Cu3O7 are both high-Tc superconductors attaining zero resistance above 80 K. Preliminary studies indicate that Yb1−xYxBa2Cu3O7 also exhibits zero resistance above 77 K.

HIGH-Tc superconductivity in perovskite oxides of the Y-Ba-Cu-O system

Several biphasic compositions of the type Y3-xBa3+xCu6O14 show an onset of superconductivity in the 90-115K range, attaining zero resistance in the 70-85K range. The phase responsible for superconductivity in these compositions is a perovskite oxide of composition YBa2Cu3 O7. This oxide annealed in oxygen shows the onset of superconductivity at 120K and zero resistance at 87K. YBa2,Cu3O7 shows the highest Meissner effect of all oxide superconductors. The superconducting behaviour of the two perovskite oxides, Y0.95Ba1.95,Cu3O7 and Y1.05Ba1.95Cu3O7 show interesting features; a marked decrease in resistivity is observed from room temperature itself in the former oxide with zero resistance at 89K. Electron microscopy and infrared spectra of these oxides are briefly discussed.

Helix-coil stability constants for amino acids in nonaqueous solvents. Studies of random poly(-benzyl-L-glutamate-co-L-alanine) and poly(-benzyl-L-glutamate-co-L-leucine) in a mixed solvent

The host-guest technique has been applied to the determination of the helix-coil stability constants of two naturally occurring amino acids, L-alanine and L-leucine, in a nonaqueous solvent system. Random copolymers containing L-alanine and L-leucine, respectively, as guest residues and -benzyl-L-glutamate as the host residue were synthesized. The polymers were fractionated and characterized for their amino acid content, molecular weight, and helix-coil transition behavior in a dichloroacetic acid (DCA)-1,2-dichloroethane (DCE) mixture. Two types of helix-coil transitions were carried out on the copolymers: solvent-induced transitions in DCA-DCE mixtures at 25°C and thermally induced transitions in a 82:18 (wt %) DCA-DCE mixture. The thermally induced transitions were analyzed by statistical mechanical methods to determine the Zimm-Bragg parameters, and s, of the guest residues. The experimental data indicate that, in the nonaqueous solvent, the L-alanine residue stabilizes the -helical conformation more than the L-leucine residue does. This is in contrast to their behavior in aqueous solution, where the reverse is true. The implications of this finding for the analysis of helical structures in globular proteins are discussed.

X-ray absorption spectra of niobium dichalcogenides and their first-row transition-metal intercalates

X-ra!. K-absorption spectra of niobium in niobium dichalcogenides. namely NbS, and NbSe, and their first-row transition-metal intercalates Mi P 3N bSz (M = Cr. Mn. Fe. Co. Ni) and Ml#,NbSe2 (M = Fe. CO). have been measured together with those in niobium metal. The spectra of these materials are \er? similar to one another. They reflect the transitions to the partially filled niobium d band with some p character. A bariety of x-ray absorption nearedge structures (XASES) associated with the K edges of intercalated atoms are also presented and discussed.

Evolution of three-dimensional character across the Lan+1NinO3n+1 homologous series with increase in n1

Electrical and magnetic properties of La3Ni2O7 and La4Ni3O10 have been investigated in comparison with those of La2NiO4, LaNiO3, and LaSrNiO4. The results suggest an increasing 3-dimensional character across the homologous series Lan+1NinO3n+1 with increase in n. Accordingly, the electrical resistivity decreases in the order La3Ni2O7, La4Ni3O10, and LaNiO3 and this trend is suggested to be related to the percolation threshold. Magnetic properties of these oxides also show some interesting trends across the series.