arXiv subjects

Chen Zeng

Publications and source records attributed to Chen Zeng.

36 records · Page 2Linked to original sources

arXiv2021

deGraphCS: Embedding Variable-based Flow Graph for Neural Code Search

With the rapid increase in the amount of public code repositories, developers maintain a great desire to retrieve precise code snippets by using natural language. Despite existing deep learning based approaches(e.g., DeepCS and MMAN) have provided the end-to-end solutions (i.e., accepts natural language as queries and shows related code fragments retrieved directly from code corpus), the accuracy of code search in the large-scale repositories is still limited by the code representation (e.g., AST) and modeling (e.g., directly fusing the features in the attention stage). In this paper, we propose a novel learnable deep Graph for Code Search (calleddeGraphCS), to transfer source code into variable-based flow graphs based on the intermediate representation technique, which can model code semantics more precisely compared to process the code as text directly or use the syntactic tree representation. Furthermore, we propose a well-designed graph optimization mechanism to refine the code representation, and apply an improved gated graph neural network to model variable-based flow graphs. To evaluate the effectiveness of deGraphCS, we collect a large-scale dataset from GitHub containing 41,152 code snippets written in C language, and reproduce several typical deep code search methods for comparison. Besides, we design a qualitative user study to verify the practical value of our approach. The experimental results have shown that deGraphCS can achieve state-of-the-art performances, and accurately retrieve code snippets satisfying the needs of the users.

cs.SE↗

arXiv2021

DepOwl: Detecting Dependency Bugs to Prevent Compatibility Failures

Applications depend on libraries to avoid reinventing the wheel. Libraries may have incompatible changes during evolving. As a result, applications will suffer from compatibility failures. There has been much research on addressing detecting incompatible changes in libraries, or helping applications co-evolve with the libraries. The existing solution helps the latest application version work well against the latest library version as an afterthought. However, end users have already been suffering from the failures and have to wait for new versions. In this paper, we propose DepOwl, a practical tool helping users prevent compatibility failures. The key idea is to avoid using incompatible versions from the very beginning. We evaluated DepOwl on 38 known compatibility failures from StackOverflow, and DepOwl can prevent 32 of them. We also evaluated DepOwl using the software repository shipped with Ubuntu-19.10. DepOwl detected 77 unknown dependency bugs, which may lead to compatibility failures.

cs.SE↗

arXiv2020

Deep Learning based Multi-Modal Sensing for Tracking and State Extraction of Small Quadcopters

This paper proposes a multi-sensor based approach to detect, track, and localize a quadcopter unmanned aerial vehicle (UAV). Specifically, a pipeline is developed to process monocular RGB and thermal video (captured from a fixed platform) to detect and track the UAV in our FoV. Subsequently, a 2D planar lidar is used to allow conversion of pixel data to actual distance measurements, and thereby enable localization of the UAV in global coordinates. The monocular data is processed through a deep learning-based object detection method that computes an initial bounding box for the UAV. The thermal data is processed through a thresholding and Kalman filter approach to detect and track the bounding box. Training and testing data are prepared by combining a set of original experiments conducted in a motion capture environment and publicly available UAV image data. The new pipeline compares favorably to existing methods and demonstrates promising tracking and localization capacity of sample experiments.

cs.CV↗

arXiv2020

Evolution of the Earth's Magnetosheath Turbulence: A statistical study based on MMS observations

Composed of shocked solar wind, the Earth's magnetosheath serves as a natural laboratory to study the transition of turbulence from low Alfv{é}n Mach number, $M_\mathrm{A}$, to high $M_\mathrm{A}$. The simultaneous observations of magnetic field and plasma moments with unprecedented high temporal resolution provided by NASA's \textit{Magnetospheric Multiscale} Mission enable us to study the magnetosheath turbulence at both magnetohydrodynamics (MHD) and sub-ion scales. Based on 1841 burst-mode segments of MMS-1 from 2015/09 to 2019/06, comprehensive patterns of the spatial evolution of magnetosheath turbulences are obtained: (1) from the sub-solar region to the flanks, $M_\mathrm{A}$ increases from $<$ 1 to $>$ 5. At MHD scales, the spectral indices of the magnetic-field and velocity spectra present a positive and negative correlation with $M_\mathrm{A}$. However, no obvious correlations between the spectral indices and $M_\mathrm{A}$ are found at sub-ion scales. (2) from the bow shock to the magnetopause, the turbulent sonic Mach number, $M_{\mathrm{turb}}$, generally decreases from $>$ 0.4 to $<$ 0.1. All spectra steepen at MHD scales and flatten at sub-ion scales, representing a positive/negative correlations with $M_\mathrm{turb}$. The break frequency increases by 0.1 Hz when approaching the magnetopause for the magnetic-field and velocity spectra, while it remains at 0.3 Hz for the density spectra. (3) In spite of some differences, similar results are found for the quasi-parallel and quasi-perpendicular magnetosheath. In addition, the spatial evolution of magnetosheath turbulence is found to be independent of the upstream solar wind conditions, e.g., the Z-component of the interplanetary magnetic field and the solar wind speed.

physics.space-ph↗

arXiv2019

Hour-scale persistent negative anomaly of atmospheric electrostatic field near the epicenter before earthquake

Although earthquake prediction is a big challenge in the world, some simple observational tools can capture many physical signals and demonstrate that an earthquake (EQ) may be forthcoming in short period. Many researchers have studied the significant variation of atmospheric electrostatic field related to the forthcoming earthquake. However, until now, there is not a compelling physical mechanism which can explain why atmospheric electrostatic abnormal signal could appear just before an earthquake. Here we present a precursor signal and propose a brief physical interpretation. Under fair air conditions, if the near-surface atmospheric electrostatic field Ez (oriented down when it is positive) presents a very stable negative anomaly (from -100 V/m to -5000 V/m), it will forebode that an earthquake (seismicity from 3-8) would take place in the next several to tens of hours within a distance less than 100 km. We name this prediction technique as "DC Ez Determination"(DED). In addition, the mechanism of such abnormal quasi-static electric field before a forthcoming earthquake has been proposed here: (1) Radon gas releases from the rock clefts near the epicenter during seismogenic process. (2) α particles are produced due to the radioactive decay of Radon gas in the air. (3) α particle ionizing radiation creates more positive particles in the air, which is much more effective than that β and γ particles produced during Radon radioactive decay. (4) The new positive particles change formal positive atmospheric electric field (Ez) into stably negative near the earth-surface. (5) The closer the instrument is to the epicenter, the more likely it is to observe the negative Ez signal related to the earthquake. It is recommended to establish an instrument network to capture the reliable precursor and make some warnings before the earthquake disaster.

physics.geo-ph↗

arXiv2017

Direct Information Reweighted by Contact Templates: Improved RNA Contact Prediction by Combining Structural Features

It is acknowledged that co-evolutionary nucleotide-nucleotide interactions are essential for RNA structures and functions. Currently, direct coupling analysis (DCA) infers nucleotide contacts in a sequence from its homologous sequence alignment across different species. DCA and similar approaches that use sequence information alone usually yield a low accuracy, especially when the available homologous sequences are limited. Here we present a new method that incorporates a Restricted Boltzmann Machine (RBM) to augment the information on sequence co-variations with structural patterns in contact inference. We thus name our method DIRECT that stands for Direct Information REweighted by Contact Templates. Benchmark tests demonstrate that DIRECT produces a substantial enhancement of 13% in accuracy on average for contact prediction in comparison to the traditional DCA. These results suggest that DIRECT could be used for improving predictions of RNA tertiary structures and functions. The source codes and dataset of DIRECT are available at http:// http://zhao.phy.ccnu.edu.cn:8122/DIRECT/index.html.

q-bio.BM↗

arXiv2003

Test of Replica Theory: Thermodynamics of 2D Model Systems with Quenched Disorder

We study the statistics of thermodynamic quantities in two related systems with quenched disorder: A (1+1)-dimensional planar lattice of elastic lines in a random potential and the 2-dimensional random bond dimer model. The first system is examined by a replica-symmetric Bethe ansatz (RBA) while the latter is studied numerically by a polynomial algorithm which circumvents slow glassy dynamics. We establish a mapping of the two models which allows for a detailed comparison of RBA predictions and simulations. Over a wide range of disorder strength, the effective lattice stiffness and cumulants of various thermodynamic quantities in both approaches are found to agree excellently. Our comparison provides, for the first time, a detailed quantitative confirmation of the replica approach and renders the planar line lattice a unique testing ground for concepts in random systems.

cond-mat.dis-nn↗

arXiv2001

Identifying Proteins of High Designability via Surface-Exposure Patterns

Using an off-lattice model, we fully enumerate folded conformations of polypeptide chains of up to N = 19 monomers. Structures are found to differ markedly in designability, defined as the number of sequences with that structure as a unique lowest-energy conformation. We find that designability is closely correlated with the pattern of surface exposure of the folded structure. For longer chains, complete enumeration of structures is impractical. Instead, structures can be randomly sampled, and relative designability estimated either from designability within the random sample, or directly from surface-exposure pattern. We compare the surface-exposure patterns of those structures identified as highly designable to the patterns of naturally occurring proteins.

cond-mat.soft↗

arXiv1999

Universal Susceptibility Variations in 1+1 Dimensional Vortex Glass

We model a planar array of fluxlines as a discrete solid-on-solid model with quenched disorder. Simulations at finite temperatures are made possible by a new algorithm which circumvents the slow glassy dynamics encountered by traditional Metropolis Monte Carlo algorithms. Numerical results on magnetic susceptibility variations support analytic predictions.

cond-mat.dis-nn↗

arXiv1999

Thermodynamics of Mesoscopic Vortex Systems in 1+1 Dimensions

The thermodynamics of a disordered planar vortex array is studied numerically using a new polynomial algorithm which circumvents slow glassy dynamics. Close to the glass transition, the anomalous vortex displacement is found to agree well with the prediction of the renormalization-group theory. Interesting behaviors such as the universal statistics of magnetic susceptibility variations are observed in both the dense and dilute regimes of this mesoscopic vortex system.

cond-mat.dis-nn↗

arXiv1999

Simulation of the Zero Temperature Behavior of a 3-Dimensional Elastic Medium

We have performed numerical simulation of a 3-dimensional elastic medium, with scalar displacements, subject to quenched disorder. We applied an efficient combinatorial optimization algorithm to generate exact ground states for an interface representation. Our results indicate that this Bragg glass is characterized by power law divergences in the structure factor $S(k)\sim A k^{-3}$. We have found numerically consistent values of the coefficient $A$ for two lattice discretizations of the medium, supporting universality for $A$ in the isotropic systems considered here. We also examine the response of the ground state to the change in boundary conditions that corresponds to introducing a single dislocation loop encircling the system. Our results indicate that the domain walls formed by this change are highly convoluted, with a fractal dimension $d_f=2.60(5)$. We also discuss the implications of the domain wall energetics for the stability of the Bragg glass phase. As in other disordered systems, perturbations of relative strength $δ$ introduce a new length scale $L^* \sim δ^{-1/ζ}$ beyond which the perturbed ground state becomes uncorrelated with the reference (unperturbed) ground state. We have performed scaling analysis of the response of the ground state to the perturbations and obtain $ζ= 0.385(40)$. This value is consistent with the scaling relation $ζ=d_f/2- θ$, where $θ$ characterizes the scaling of the energy fluctuations of low energy excitations.

cond-mat.dis-nn↗

arXiv1998

Application of Polynomial Algorithms to a Random Elastic Medium

A randomly pinned elastic medium in two dimensions is modeled by a disordered fully-packed loop model. The energetics of disorder-induced dislocations is studied using exact and polynomial algorithms from combinatorial optimization. Dislocations are found to become unbound at large scale, and the elastic phase is thus unstable giving evidence for the absence of a Bragg glass in two dimensions.

cond-mat.dis-nn↗

arXiv1998

Absence of Two-Dimensional Bragg Glasses

The stability to dislocations of the elastic phase, or ``Bragg glass'', of a randomly pinned elastic medium in two dimensions is studied using the minimum-cost-flow algorithm for a disordered fully-packed loop model. The elastic phase is found to be unstable to dislocations due to the quenched disorder. The energetics of dislocations are discussed within the framework of renormalization group predictions as well as in terms of a domain wall picture.

cond-mat.dis-nn↗

arXiv1998

Ab Initio Simulation of the Nodal Surfaces of Heisenberg Antiferromagnets

The spin-half Heisenberg antiferromagnet (HAF) on the square and triangular lattices is studied using the coupled cluster method (CCM) technique of quantum many-body theory. The phase relations between different expansion coefficients of the ground-state wave function in an Ising basis for the square lattice HAF is exactly known via the Marshall-Peierls sign rule, although no equivalent sign rule has yet been obtained for the triangular lattice HAF. Here the CCM is used to give accurate estimates for the Ising-expansion coefficients for these systems, and CCM results are noted to be fully consistent with the Marshall-Peierls sign rule for the square lattice case. For the triangular lattice HAF, a heuristic rule is presented which fits our CCM results for the Ising-expansion coefficients of states which correspond to two-body excitations with respect to the reference state. It is also seen that Ising-expansion coefficients which describe localised, $m$-body excitations with respect to the reference state are found to be highly converged, and from this result we infer that the nodal surface of the triangular lattice HAF is being accurately modeled. Using these results, we are able to make suggestions regarding possible extensions of existing quantum Monte Carlo simulations for the triangular lattice HAF.

cond-mat.str-el↗

arXiv1997

Statistical Topography of Glassy Interfaces

Statistical topography of two-dimensional interfaces in the presence of quenched disorder is studied utilizing combinatorial optimization algorithms. Finite-size scaling is used to measure geometrical exponents associated with contour loops and fully packed loops. We find that contour-loop exponents depend on the type of disorder (periodic ``vs'' non-periodic) and they satisfy scaling relations characteristic of self-affine rough surfaces. Fully packed loops on the other hand are unaffected by disorder with geometrical exponents that take on their pure values.

cond-mat.dis-nn↗

arXiv1996

An efficient implementation of high-order coupled-cluster techniques applied to quantum magnets

We illustrate how the systematic inclusion of multi-spin correlations of the quantum spin-lattice systems can be efficiently implemented within the framework of the coupled-cluster method by examining the ground-state properties of both the square-lattice and the frustrated triangular-lattice quantum antiferromagnets. Various physical quantities including the ground-state energy, the anisotropy susceptibility, and the sublattice magnetisation are calculated and compared with those obtained from such other methods as series expansions and quantum Monte Carlo simulations.

cond-mat.stat-mech↗

arXiv1996

Ground-State Roughness of the Disordered Substrate and Flux Line in d=2

We apply optimization algorithms to the problem of finding ground states for crystalline surfaces and flux lines arrays in presence of disorder. The algorithms provide ground states in polynomial time, which provides for a more precise study of the interface widths than from Monte Carlo simulations at finite temperature. Using $d=2$ systems up to size $420^2$, with a minimum of $2 \times 10^3$ realizations at each size, we find very strong evidence for a $\ln^2(L)$ super-rough state at low temperatures.

cond-mat↗

arXiv1995

Numerical and approximate analytical results for the frustrated spin-1/2 quantum spin chain

We study the $T=0$ frustrated phase of the $1D$ quantum spin-$\frac 12$ system with nearest-neighbour and next-nearest-neighbour isotropic exchange known as the Majumdar-Ghosh Hamiltonian. We first apply the coupled-cluster method of quantum many-body theory based on a spiral model state to obtain the ground state energy and the pitch angle. These results are compared with accurate numerical results using the density matrix renormalisation group method, which also gives the correlation functions. We also investigate the periodicity of the phase using the Marshall sign criterion. We discuss particularly the behaviour close to the phase transitions at each end of the frustrated phase.

cond-mat↗