AMAshwin Vishnu Mohanan

Ashwin Vishnu Mohanan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Ashwin Vishnu Mohanan has a total of 29 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Introduction to deep learning: Carpentries-style hands-on lesson material for introducing researchers to deep learning

    by Sven A. van der Burg, Pranav Chandramouli, Anne Fouilloux, Cunliang Geng, Toby Hodges, Florian Huber, Dafne van Kuppevelt, Ashwin Vishnu Mohanan, Colin Sauze, Carsten Schnober, et al.

    This article describes a hands-on introduction to the first steps in deep learning, intended for researchers who are familiar with (non-deep) machine learning. The use of deep learning has seen a sharp increase in popularity and applicability over the last decade. While deep learning can be a useful tool for researchers from a wide range of domains, taking the first steps in the world of deep learning can be somewhat intimidating. This introduction aims to cover the fundamentals of deep learning in a practical and hands-on manner. By the end of the course, students will be able to train their first neural network and understand the subsequent steps needed to improve the model. The lesson starts by explaining the basic concepts of neural networks, and then guides learners through the different steps of a deep learning workflow. After following this lesson, learners will be able to prepare data for deep learning, implement a basic deep learning model in Python with Keras, and monitor and troubleshoot the training process. In addition, they will be able to implement and understand different layer types, such as convolutional layers and dropout layers, and apply transfer learning. We use data with permissive licenses and designed for real world use cases: • The Penguin dataset (Horst et al. (2020)) • The Weather prediction dataset (Huber et al. (2022)) • The Dollar Street Dataset (Gaviria Rojas et al. (2022)) is representative and contains accurate demographic information to ensure their robustness and fairness, especially for smaller subpopulations. Statement of Need This lesson addresses the need for an introductory lesson on deep learning that is opensource, and can be used by instructors in a workshop as well as for self-study. While generally usable, its target audience are academic researchers.


  • Vertical convection regimes in a two-dimensional rectangular cavity: Prandtl and aspect ratio dependence

    by Arman Khoubani, Ashwin Vishnu Mohanan, Pierre Augier, Jan-Bert Flór

    Vertical convection is the fluid motion that is induced by the heating and cooling of two opposed vertical boundaries of a rectangular cavity (see e.g. Wang et al., J. Fluid Mech., vol. 917, 2021, A6). We consider the linear stability of the steady two-dimensional flow reached at Rayleigh numbers of O( $10^8$ ). As a function of the Prandtl number, $Pr$ , and the height-to-width aspect ratio of the domain, $A$ , the base flow of each case is computed numerically and linear simulations are used to obtain the properties of the leading linear instability mode. Flow regimes depend on the presence of a circulation in the entire cavity, detachment of the thermal layer from the boundary or the corner regions and on the oscillation frequency relative to the natural frequency of oscillation in the stably temperature-stratified interior, allowing for the presence of internal waves or not. Accordingly, the regime is called slow or fast, respectively. Either the global circulation or internal waves in the interior may couple the top and bottom buoyancy currents, while their absence implies asymmetry in their perturbation amplitude. Six flow regimes are found in the range of $0.1 \leq Pr \leq 4$ and $0.5 \leq A \leq 2$ . For $Pr \lessapprox 0.4$ and $A>1$ , the base flow is driven by a large circulation in the entire cavity. For $Pr \gtrapprox 0.7$ , the thermal boundary layers are thin and the instability is driven by the motion along the wall and the detached boundary layer. A transition between these regimes is marked by a dramatic change in oscillation frequency at $Pr = 0.55 \pm 0.15$ and $A <2$ .


  • Open Access

    Snek5000: a new Python framework for Nek5000

    by Ashwin Vishnu Mohanan, Arman Khoubani, Pierre Augier


  • Reducing the ecological impact of computing through education and Python compilers

    by Pierre Augier, Carl Friedrich Bolz-Tereick, Serge Guelton, Ashwin Vishnu Mohanan


  • Open Access

    Shallow water wave turbulence

    by Pierre Augier, Ashwin Vishnu Mohanan, Erik Lindborg

    The dynamics of irrotational shallow water wave turbulence forced at large scales and dissipated at small scales is investigated. First, we derive the shallow water analogue of the ‘four-fifths law’ of Kolmogorov turbulence for a third-order structure function involving velocity and displacement increments. Using this relation and assuming that the flow is dominated by shocks, we develop a simple model predicting that the shock amplitude scales as $(\unicode[STIX]{x1D716}d)^{1/3}$ , where $\unicode[STIX]{x1D716}$ is the mean dissipation rate and $d$ the mean distance between the shocks, and that the $p$ th-order displacement and velocity structure functions scale as $(\unicode[STIX]{x1D716}d)^{p/3}r/d$ , where $r$ is the separation. Then we carry out a series of forced simulations with resolutions up to $7680^{2}$ , varying the Froude number, $F_{f}=(\unicode[STIX]{x1D716}L_{f})^{1/3}/c$ , where $L_{f}$ is the forcing length scale and $c$ is the wave speed. In all simulations a stationary state is reached in which there is a constant spectral energy flux and equipartition between kinetic and potential energy in the constant flux range. The third-order structure function relation is satisfied with a high degree of accuracy. Mean energy is found to scale approximately as $E\sim \sqrt{\unicode[STIX]{x1D716}L_{f}c}$ , and is also dependent on resolution, indicating that shallow water wave turbulence does not fit into the paradigm of a Richardson–Kolmogorov cascade. In all simulations shocks develop, displayed as long thin bands of negative divergence in flow visualisations. The mean distance between the shocks is found to scale as $d\sim F_{f}^{1/2}L_{f}$ . Structure functions of second and higher order are found to scale in good agreement with the model. We conclude that in the weak limit, $F_{f}\rightarrow 0$ , shocks will become denser and weaker and finally disappear for a finite Reynolds number. On the other hand, for a given $F_{f}$ , no matter how small, shocks will prevail if the Reynolds number is sufficiently large.


  • Open Access

    FluidSim: Modular, Object-Oriented Python Package for High-Performance CFD Simulations

    by Ashwin Vishnu Mohanan, Cyrille Bonamy, Miguel Calpe Linares, Pierre Augier


  • Open Access

    FluidFFT: Common API (C++ and Python) for Fast Fourier Transform HPC Libraries

    by Ashwin Vishnu Mohanan, Cyrille Bonamy, Pierre Augier

    The Python package fluidfft provides a common Python API for performing Fast Fourier Transforms (FFT) in sequential, in parallel and on GPU with different FFT libraries (FFTW, P3DFFT, PFFT, cuFFT). fluidfft is a comprehensive FFT framework which allows Python users to easily and efficiently perform FFT and the associated tasks, such as computing linear operators and energy spectra. We describe the architecture of the package composed of C++ and Cython FFT classes, Python "operator" classes and Pythran functions. The package supplies utilities to easily test itself and benchmark the different FFT solutions for a particular case and on a particular machine. We present a performance scaling analysis on three different computing clusters and a microbenchmark showing that fluidfft is an interesting solution to write efficient Python applications using FFT.


  • A two-dimensional toy model for geophysical turbulence

    by Erik Lindborg, Ashwin Vishnu Mohanan

    A toy model for large scale geophysical turbulence is constructed by making two modifications of the shallow water model. Unlike the shallow water model, the toy model has a quadratic expression for total energy, which is the sum of Available Potential Energy (APE) and Kinetic Energy (KE). More importantly, in contrast to the shallow water model, the toy model does not produce any shocks. Three numerical simulations with different forcing are presented and compared with the simulation of a full General Circulation Model (GCM). The energy which is injected cascades in a similar way as in the GCM. First, some of the energy is converted from APE to KE at large scales. The wave field then undergoes a forward energy cascade displaying shallow spectra, close to k−5/3, for both APE and KE, while the vortical field either displays a k−3-spectrum or a more shallow spectrum, close to k−5/3, depending on the forcing. In a simulation with medium forcing wave number, some of the energy which is converted from APE to KE undergoes an inverse energy cascade which is produced by nonlinear interactions only involving the rotational component of the velocity field. The inverse energy cascade builds up a vortical field at larger scales than the forcing scale. At these scales, coherent vortices emerge with a strong dominance of anticyclonic vortices. The relevance of the simulation results to the dynamics of the atmosphere is discussed as in possible continuations of the investigation.


  • KdV Equation and Computations of Solitons: Nonlinear Error Dynamics

    by V. M. Ashwin, K. Saurabh, M. Sriramkrishnan, P. M. Bagade, M. K. Parvathi, Tapan K. Sengupta