Clone-advisor: recommending code tokens and clone methods with deep learning and information retrieval

Software developers frequently reuse source code from repositories as it saves development time and effort. Code clones (similar code fragments) accumulated in these repositories represent often repeated functionalities and are candidates for reuse in an exploratory or rapid development. To facilitate code clone reuse, we previously presented DeepClone, a novel deep learning approach for modeling code clones along with non-cloned code to predict the next set of tokens (possibly a complete clone method body) based on the code written so far. The probabilistic nature of language modeling, however, can lead to code output with minor syntax or logic errors. To resolve this, we propose a novel approach called Clone-Advisor. We apply an information retrieval technique on top of DeepClone output to recommend real clone methods closely matching the predicted clone method, thus improving the original output by DeepClone. In this paper we have discussed and refined our previous work on DeepClone in much more detail. Moreover, we have quantitatively evaluated the performance and effectiveness of Clone-Advisor in clone method recommendation.

Supplemental Material: Recalibrating the Devonian time scale: A new method for integrating radioisotopic and astrochronologic ages in a Bayesian framework

Laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) methods and results, additional age-depth modeling methods, and description of modeling code; LA-ICP-MS data for Hercules I K-bentonite and age-depth model inputs and results; R script for executing age-depth modeling procedure; and R script for anchoring floating astrochronology durations.

Supplemental Material: Recalibrating the Devonian time scale: A new method for integrating radioisotopic and astrochronologic ages in a Bayesian framework

Laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) methods and results, additional age-depth modeling methods, and description of modeling code; LA-ICP-MS data for Hercules I K-bentonite and age-depth model inputs and results; R script for executing age-depth modeling procedure; and R script for anchoring floating astrochronology durations.

High-Energy Mechanical Milling-Driven Reamorphization in Glassy Arsenic Monoselenide: On the Path of Tailoring Special Molecular-Network Glasses

The impact of high-energy milling on glassy arsenic monoselenide g-AsSe is studied with X-ray diffraction applied to diffuse peak-halos proper to intermediate- and extended-range ordering revealed in first and second sharp diffraction peaks (FSDP and SSDP). A straightforward interpretation of this effect is developed within the modified microcrystalline approach, treating “amorphous” halos as a superposition of the broadened Bragg diffraction reflexes from remnants of some inter-planar correlations, supplemented by the Ehrenfest diffraction reflexes from most prominent inter-molecular and inter-atomic correlations belonging to these quasi-crystalline remnants. Under nanomilling, the cage-like As4Se4 molecules are merely destroyed in g-AsSe, facilitating a more polymerized chain-like network. The effect of nanomilling-driven molecular-to-network reamorphization results in a fragmentation impact on the correlation length of FSDP-responsible entities (due to an increase in the FSDP width and position). A breakdown in intermediate-range ordering is accompanied by changes in extended-range ordering due to the high-angular shift and broadening of the SSDP. A breakdown in the intermediate-range order is revealed in the destruction of most distant inter-atomic correlations, which belong to remnants of some quasi-crystalline planes, whereas the longer correlations dominate in the extended-range order. The microstructure scenarios of milling-driven reamorphization originated from the As4Se4 molecule, and its network derivatives are identified with an ab initio quantum-chemical cluster modeling code (CINCA).

Dust and minerals in the HR 8799 atmospheres: JWST MIRI prediction

<p>We are learning rapidly about the gas composition of exoplanet atmospheres, but know almost nothing about their solid composition. The upcoming James Webb Space Telescope will radically change this! The HR8799 exoplanetary system is a perfect candidate because it provides us with a unique opportunity of simultaneously measuring mineral clouds and refractory element composition of its four gas giant atmospheres. The HR8799 system is very young and additionally contains two particle belts. The giant planets are predicted to be bombarded with material from the belts, analogous to the Late Heavy Bombardment. Signatures of this bombardment, such as mineral clouds and refractory element composition, might be observable in their atmospheres. JWST MIRI will allow to characterise these exoplanets in the mid-infrared thermal regime (5-28 μm) which is not possible from the ground. We use the ARtful modeling Code for exoplanet Science (ARCiS) to calculate the mid-infrared spectra of planets HR 8799 b, c, d and e, and we simulate MIRI spectroscopic observations. Besides the dust features, we also expect to identify narrower gas features from molecular species such as CO<sub>2</sub>, H<sub>2</sub>O, HCN, C<sub>2</sub>H<sub>2</sub> etc.</p>

HST WFC3 G141 data analysis: exploring the transition from Super-Earth to Sub-Neptune

<div> <div> <div> <p>Exoplanets with size between the Earth and Neptune (1-4R⊕) do not have any equivalent in our Solar System and remain challenging to characterize. Yet, there are ubiquitous in the Galaxy and Fulton et al. (2017) showed that their distribution (number of planets per star vs radius) is bimodal highlighting a gap in the number of planets around 1.7R⊕. Planets with a radius below 1.7R⊕ are thought to be mostly rocky planets, and called Super-Earth, above this limit planets are most likely made of gas and called Sub-Neptune. We made use of the available data from the Hubble Space Telescope in Near-Infrared (HST WFC3 G141) and gathered 18 transmission spectra of planets with size below 6 R⊕ to study the transition between rocky and gaseous planets. First, we used TauREx3 (Al-Refaie et al. 2019), a Bayesian retrieval code, to rule out atmospheric scenarios. We proved that a primary clear atmosphere dominated by Hydrogen and Helium is rejected with more than 3σ for a large majority of planets in the sample. Then, we measured the amplitude of the spectra in the water absorption band (around 1.4μm) and compared observational values to simulated ones using a self-consistent modeling code ExoREM (Baudino et al. 2015; Charnay et al. 2018). We explored the connection between the water absorption amplitudes and the temperature by setting the stellar and planetary parameters to those of HD 3167 c (2.7 R⊕, 8.33 M⊕) and trying different metallicities (1, 10, 100 and 1000 x solar), cloud compositions and temperatures (300-1200K).</p> </div> </div> </div>

Cluster Modeling of Network-Forming Amorphization Pathways in AsxS100−x Arsenicals (50 ≤ x ≤ 57) Diven by Nanomilling

Complete hierarchy of network amorphization scenarios initiated in AsxS100-x nanoarsenicals within As4S4-As4S3 cut-Sect. (50 ≤ x ≤ 57) is reconstructed employing materials-computational approach based on ab-initio quantum-chemical modeling code (CINCA). Under nanostructurization due to high-energy mechanical milling, the inter-crystalline transformations to nanoscopic β-As4S4 phase accompanied by appearance of covalent-network amorphous matrix are activated. General amorphization trend under nanomilling obeys tending from molecular cage-like structures to optimally-constrained covalent-bonded networks compositionally invariant with parent arsenical. The contribution of amorphization paths in nanoarsenicals is defined by their chemistry with higher molecular-to-network barriers proper to As4S3-rich alloys. The generated amorphous phase is intrinsically decomposed, possessing double-Tg relaxation due to stoichiometric (x = 40) and non-stoichiometric (x > 40) sub-networks, which are built of AsS3/2 pyramids and As-rich arrangement keeping (i) two separated As-As bonds derived from realgar-type molecules, (ii) two neighboring As-As bonds derived from pararealgar-type molecules or (iii) three neighboring As-As bonds in triangle-like geometry derived from dimorphite-type molecules. Compositional invariance of nanoamorphous phase is ensured by growing sequence of network-forming clusters with average coordination numbers Z in the row (As2S4/2,Z = 2.50) – (As3S5/2, Z = 2.55) – (As3S3/2, Z = 2.67). Diversity of main molecular-to-network amorphizing pathways in nanoarsenicals is reflected on the unified potential energy landscape specified for boundary As4S4 and As4S3 components.

Spatio-temporal patterns of Pyrenean exhumation revealed by inverse thermo-kinematic modeling of a large thermochronologic data set

Large thermochronologic data sets enable orogen-scale investigations into spatio-temporal patterns of erosion and deformation. We present the results of a thermo-kinematic modeling study that examines large-scale controls on spatio-temporal variations in exhumation as recorded by multiple low-temperature thermochronometers in the Pyrenees mountains (France/Spain). Using 264 compiled cooling ages spanning ~200 km of the orogen, a recent model for its topographic evolution, and the thermo-kinematic modeling code Pecube, we evaluated two models for Axial Zone (AZ) exhumation: (1) thrust sheet–controlled (north-south) exhumation, and (2) along-strike (east-west) variable exhumation. We also measured the degree to which spatially variable post-orogenic erosion influenced the cooling ages. We found the best fit for a model of along-strike variable exhumation. In the eastern AZ, rock uplift rates peak at ≥1 mm/yr between 40 and 30 Ma, whereas in the western AZ, they peak between 30 and 20 Ma. The amount of post-orogenic (<20 Ma) erosion increases from <1.0 km in the eastern Pyrenees to >2.5 km in the west. The data reveal a pattern of exhumation that is primarily controlled by structural inheritance, with ancillary patterns reflecting growth and erosion of the antiformal stack and post-orogenic surface processes.