Software and Code

Permanent link for this collectionhttps://hdl.handle.net/2022/21617

Browse

Recent Submissions

Now showing 1 - 5 of 5
  • Item
    Jetstream Unofficial Addon for Atmosphere VM(s)
    (Zenodo, 2017-08-16) Roy, Samapriya
    Jetstream provides researchers and students with a unique approach to cloud computing and data analysis on demand. Moving away from a job submission model which is more pervasive on High-Performance Cluster computing this allows users to create user-friendly installs of “virtual machines” which can then be custom installed with softwares and packages and even imaged further to make science replicable. The system in theory also allows a BYoOS (Bring your own Operating System) designed to further allow more flexibility while sharing the backbone of high performance compute and cloud storage infrastructure. The idea of imaging a system allows the user to not only share their data or analysis but their entire setup that allows them to run the analysis making it a plug and play model in terms of research replication and transference. As the NSF project proceeds further it allows users to apply for grants or allocation time on the machines and indeed renew these free of cost for researchers and students.While working on my projects I realized it would be easier to query the number of instances and volumes running on my account, my burn rate and left over allocation over time. And something useful was the possibility to perform system actions such as shutdown(stop), restart, start, reboot among others programmatically. It is with this idea that I approach the API backend at JetStream and built just a few tools to allows users to perform these actions without the need to log into their web browsers and allows the user to call upon instance actions through a command line interface (CLI).The tool creates a payload which carries the then creates get or post requests allowing us to perform these actions.
  • Item
    Google Earth Engine Asset Manager with Addons
    (Zenodo, 2017-06-19) Roy, Samapriya
    Google Earth Engine(GEE) is a cloud-based image analysis and geospatial system designed to perform complex functions and includes backend for both storage and compute processes. The benefit of this system is that it is constantly updated with open source datasets that can be hosted on their servers(including but not limited to Landsat and Sentinel Archives, along with other satellite series). The array of possible tools that allows the users to perform a distributed computing request is unique and made available to those wanting to utilize this massive computing resource. There are two ways to communicate with GEE which are JavaScript based interface and a python client based command line and interaction backend. The command line tool is also very useful to perform some additional tasks such as query a user's own asset, delete , create assets, and perform on demand queries like asset size among other things. Though these tools are built with some consistent outputs this package or add-on was developed to enhance these functionalities. The additional tools include functionality for moving and copying assets, cleaning folders, querying tasks apart from uploading and deleting assets from collections. The tool also includes batch upload tools adapted from an earlier work designed to upload assets onto GEE and this tool adds functionally to that along with being a bundle of tools which can be easily used by Earth Engine users with a command line environment. The new release also contains options to change collection properties, query and generate reports of all task running, and to copy collection from one user to another or from one collection to another. It is developed case by case basis to include more features in the future as it becomes available or as need arises.
  • Item
    ArcticDEM Batch Pipeline
    (Zenodo, 2017-08-12) Roy, Samapriya
    ArcticDEM project was a joint project supported by both the National Geospatial-Intelligence Agency(NGA) and the National Science Foundation(NSF) with the idea of creating a high resolution and high-quality digital surface model(DSM). The product is distributed free of cost as time-dependent DEM strips and is hosted as https links that a user can use to download each strip. The created product is a 2-by-2 meter elevation cells over an over of over 20 million square kilometers and uses digital globe stereo imagery to create these high-resolution DSM(s). The method used for the 2m derivate is Surface Extraction with TIN-based Search-space Minimization(SETSM). With this in mind and with the potential applications of using these toolsets there was a need to batch download the DEM files for your area of interest and to be able to extract, clean and process metadata. In all fairness, this tool has a motive of extending this as an input to Google Earth Engine(GEE) and hence the last tool which is the metadata parser is designed to create a metadata manifest in a CSV file which GEE can understand and associate during asset upload.
  • Item
    Planet to Google Earth Engine Pipeline (Command Line Interface)
    (Zenodo, 2017-08-13) Roy, Samapriya
    While moving between assets from Planet Inc and Google Earth Engine it was imperative to create a pipeline that allows for easy transitions between the two service end points and this tool is designed to act as a step by step process chain from Planet Assets to batch upload and modification within the Google Earth Engine environment. The ambition is apart from helping user with batch actions on assets along with interacting and extending capabilities of existing GEE CLI. It is developed case by case basis to include more features in the future as it becomes available or as need arises. Planet assets were used as a potential platform test-bed and the tool now includes manifests to handle additional assets such as Digital Globe, Arctic-DEM and other assets will be added as metadata information becomes available. This release also contains a windows installer which bypasses the need for you to have admin permission, it does however require you to have python in the system path.
  • Item
    Planet-Pipeline-GUI
    (Zenodo, 2017-08-17) Roy, Samapriya
    As satellite data becomes more ubiquitous we are moving into an era where access to said data is handled by (Application Programming Interface) API end points and through web requests. Planet Labs has 100(s) of doves or nano satellites which they are using to create near daily satellite data for everyplace on earth. Planet releases a lot of its data for humanitarian as well as research purpose under education licensing. However the volume of data can sometimes becomes overwhelming to handle without being able to batch download and process imagery. A graphical user interface(GUI) was designed to wrap around a python based command line interface(CLI) for users to easily interact and batch download satellite data from Planet Labs. This tool allows the user to easily work with Planet's python API without any prerequisite knowledge of python or programming and to batch download data. It does however require you to have python in the system path. The tool is designed to reach those who have constrained technical backgrounds and/or resource availability while maintaining a robust back end and functionality.