{"id":1191,"date":"2013-05-31T23:33:36","date_gmt":"2013-05-31T23:33:36","guid":{"rendered":"https:\/\/notebooks.dataone.org\/?p=1191"},"modified":"2013-05-31T23:33:36","modified_gmt":"2013-05-31T23:33:36","slug":"start-the-development-of-model-comparison-package","status":"publish","type":"post","link":"https:\/\/notebooks.dataone.org\/prov-model\/start-the-development-of-model-comparison-package\/","title":{"rendered":"Start the development of model comparison package"},"content":{"rendered":"
This is the first week of my DataONE summer intern. I had a phone meeting with my mentor Bob and Yaxing in Oak Ridge National Lab on Tuesday. They introduced the work that has been done in the past and the goal we want to achieve for this summer. Basically, the goal is to develop a package that supports climate model\/data inter-comparison based on Vistrails\/UV-CDAT and finally link the package with DataONE infrastructure. During the meeting, we also decide to have formal phone meetings regularly. As we need communication quite often on the package development. We decided to use Google Hangout for instant messaging and Dropbox for file sharing.<\/p>\n
We will have a teleconference with some earth scientist from different institutions in the US next week to solicit their thoughts and needs on model\/data inter-comparison. Before the meeting, my mentors suggest me to start from implementing some simple modules that they currently have in mind, such as unit conversion, regrid, spatial subsetting and mosaicking for climate data which are recorded in NetCDF file format. The purposes of implementing those modules are manifold: to set up the environment, to understand the data format (NetCDF) that I’ll deal with most this summer and to understand how to develop user package in Vistrails environment.<\/p>\n
In terms of module development, I’ve already made the following progress during the first week:<\/p>\n
1. Current Vistrials\/UV-CDAT modules cannot preserve meta data information in NetCDF files very well. Some important information is lost during data processing and visualization process. By looking closely at the sources codes of UV-CDAT, I found where the problem is. I’m still discussing with my mentors on how to deal with this issue. Hopefully we can come up with a good solution.<\/p>\n
2. After I researched on the developer documents and source codes of Vistrials, I had a better understanding of how to extend its current functions. I created a user package called \u201cUVCDAT IMIF\u201d in Vistrails. That will be our starting points. All the following functions will be put into the package.<\/p>\n
3. As suggested by my mentors, I started to implement some basic modules within the \u201cUVCDAT IMIF\u201d user package. I finished the initial versions of \u201cunit convertor\u201d and \u201cregrid\u201d. I’ll get feedback from my mentors and improve them next week.<\/p>\n
To sum up, the first week is pretty joyful. It’s a good start.<\/p>\n
<\/p>\n","protected":false},"excerpt":{"rendered":"
This is the first week of my DataONE summer intern. I had a phone meeting with my mentor Bob and Yaxing in Oak Ridge National Lab on Tuesday. They introduced the work that has been done in the past and the goal we want to achieve for this summer. Basically, Continue reading Start the development of model comparison package<\/span>