SymPy is the computer algebra system of the SciPy stack. This talk will look at why you should care about symbolic math, some of the recent developments in SymPy, and what sorts of things you should be able to look forward to in the future.
We present an open source toolkit, UVis, that uses web 2.0 technologies and scientific python tools to provide a highly scalable, collaborative, and easy to install and use software for large climate and geospatial data analysis and visualization.
Scientific computing is not yet a part of the undergraduate physics curriculum. We have made it a focus of our junior/senior Research Methods course using unix/Linux, LaTeX, python, and SciPy. The motivation for each is provided by a series of three challenging experiments. I will discuss our methods, issues, and lessons learned.
PySIMS is Python based library of tools for analysis of time of flight secondary ion mass spectrometry (ToF-SIMS) data.
Hyperopt-sklearn is an open source hyperparameter optimization package for scikit-learn. It utilizes the power and flexibility of the hyperopt library to optimize the machine learning tools available through scikit-learn to the user's specific needs.
We present our pipeline package for the calibration and post-processing of data obtained by members of the VORTEX team using the new AGPM vector vortex coronagraph. The pipeline is written in Python 2.7 and makes extended use of the SciPy ecosystem, as well as machine-learning packages such as scikit-learn.
GNOME is a particle tracking model adapted to uses as diverse as oil spill modeling, laval transport, and tracking marine debris. This requires the model to support a wide variety of time and spatial scales, and physical processes. We will discuss the techniques used to build a system dynamic enough to accommodate this wide variety of use cases, while maximizing performance in each case.
Nikola, a full-featured static website and blog generator, interacts deeply with IPython. It is not only able to render IPython notebooks into static HTML pages, but also it supports a profound bidirectional workflow with IPython generating a new entity: The NikIPy Bridge, a full featured platform which let us setup our website/blog in the cloud and be able to write our content from everywhere.
Validated numerics is a framework for rigorous floating point computations, using interval arithmetic and directed rounding: the outcome of a calculation is guaranteed to include the correct result. ValidiPy (https://github.com/computo-fc/ValidiPy) implements core interval methods that may be immediately injected into any pre-existing code via operator overloading.
PyCube is a new visualization tool for the numerical weather prediction model on the cubed-sphere grid. The data is presented on the grid points of the cube panels and the features of a cubed-sphere coordinate system are preserved, while minimizing the loss of accuracy occurred by data processing.
The tools needed in X-ray astronomy data reduction are becoming more and more available as Python modules. I will describe the process and experience how to integrate the data reduction pipeline into Python and show advantages of spectral fitting with pyxspec. Finally I will point out modules and documentations that could be created/improved in the future.
Idealized simulations of orographic precipitation are analyzed and plotted using the scientific tools bundled with Python. A special emphasis is placed on plotting an atmospheric sounding and calculation of instability indices. The intent is to stir interest in the atmospheric science community regarding to the use of Python, which then trigger some fruitful discussions in the conference.
Pyadisi is a new open source Python package for animal locomotion research. Typical studies use high speed video cameras to record animals flying, swimming, and running. Pyadisi brings together tools already available in Python, and introduces new ones for analyzing animal locomotion video data.
Matplotlib has long been the standard for producing publication-quality plots in Python, but it's showing its age in comparison to newer tools like ggplot in R and D3js in the browser. Even so, I'd argue that matplotlib still has an important role: to this end I’ll demonstrate a number of new extensions and front-ends that I believe will keep matplotlib relevant far into the future.
A system for faster, straightforward cross-platform CPython builds across HPC, desktop, and mobile platforms with multiple build system generators and easy integration with C/C++/Fortran scientific computing libraries is described.
Using Python for analytical tools and reports with Geospatial Web-Based Portals.
Scientists use many tools, but as the scientific stack expands, so does the complexity of transitioning between tools. The OSF (http://osf.io) is a Python-based, FOSS web application that simplifies these transitions by connecting tools within a single workspace. We will describe the modular architectures that allow developers of any skill level to contribute to open science.
A python package is being constructed to efficiently and simply decode proprietary/non-standardized ascii files from disparate, commercial remote sensors into easily-used python objects. Operators of these instruments can quickly utilize the data from their instruments, which otherwise requires proprietary software or significant effort.
Rock mechanics consists of studying the behavior of natural and synthetic faults and applying this knowledge to earthquakes. The large, high resolution datasets generated by such experiments are often not shared in their raw format and many parameters analyzed by hand. Our laboratory is producing new tools and techniques to share with the community of researchers in an open source framework.
Do you hate repeating yourself? Want to know when your publication is repeating someone else? The Web of Trails project is a solution to knowledge management that empowers users to quickly find repetition of key phrases. Using syntactic indexing, as opposed to lexical techniques, this approach is capable of representing the literature using less space while providing high value results.
I will present our experience using FiPy to teach some of the thermodynamic and kinetic models used in materials science to understand microstructural evolution to audiences with different levels of experience with scientific computation, materials science, and general computer usage.
Busy scientists and engineers would love to use your Python package, but getting it installed can be a nightmare. This talk will discuss some of the tools and techniques that you can use to make it easier for users to install your Python package. We will help you choose the best method for distributing your package and suggest ways to automate the build process
PyGECoRe is a conservative remapping tool written in Python to remap between any grids in spherical geometry. PyGECoRe supports full floating-point precision and the Test Driven Development (TDD) methodology for easy maintenance.
Inverse problems haunt the nightmares of geophysics graduate students. I'll demonstrate how to conquer them using Fatiando a Terra. The new machinery in Fatiando contains many ready-to-use components and automates as much of the process as possible. You can go from zero to regularized gravity inversion with as little as 30 lines of code. I'll walk through an example to show you how.
Common tools in statistical science such as Bayesian inference, Monte Carlo, and Machine Learning impose strong assumptions on a problem set in order to yield a solution. These methods do not provide a means for testing the assumptions the methods themselves require. We have developed a framework (called 'mystic') in which assumptions in statistical science can be tested, validated, and improved.
This presentation will describe the US Geological Survey ComCat server, which hosts a comprehensive catalog of earthquake data from origins (location and magnitude) to more derived products which provide information about human impact of an earthquake (fatalities, damage, economic losses). The author will then discuss Python tools that can be used to perform complex queries of ComCat.
PESSCARA is a cloud-based infrastructure designed to promote sharing of data, metadata, algorithms, and research in medical imaging. PESSCARA consists of a content management system that stores medical images and associated metadata/tags, with a development environment based on IPython.
Much of the Big Data world is concerned with sparse and/or heterogenous data. But there are other domains where dense, homogenous data is the norm. This talk describes the issues in the provision of easy-to-use, scripting level support for "biggish" dense data (10+ GB) on standard hardware, and introduces the Biggus Python package as a way to address these challenges.
We present an open source package for Simulation and Parameter Estimation in Geophysics (SimPEG). SimPEG provides a framework for solving a variety of geophysical forward and inverse problems. We show SimPEG’s utility and extensibility through diverse examples, and discuss our design and methodologies behind creating a modular framework that promotes integration of geophysical data.
Performing modern-day scientific analysis often requires using tools from multiple domains. Here, a scientific workflow for analyzing the global shipping trade is developed using Python tools for probabilistic record linkage, fuzzy string matching, web scraping, document parsing, geodesic calculations, and raster processing to answer questions about ship movement.
Medical image analysis research requires fast prototyping, consistent data handling, and an environment where people with different expertise can easily collaborate. With these goals in mind, we utilize IPython in conjunction with a content management system to enable effortless file accessibility, quick algorithm development, and research reproducibility.
Performance analysis requires exploration and examination of data and IPython is well suited for these tasks. While IPython is an excellent tool, it is lacking features needed for large-scale development environments. In this talk, we introduce features designed to incorporate IPython into NetApp® development environment: user access controls, notebook/report viewer, and dynamic report generator.