Welcome to the Kandel Group Image Gallery.
This website allows you to access images for 70236 experimental data sets acquired from November 2001 to September 2017. This is an ongoing experiment to provide unrestricted access to our research data, with images for viewing online and raw data available for public download. The site is currently being updated and refined, but the goal is to place our entire experimental output on the web, and to update it weekly with new experimental results.
The data presented here are raw: no human effort has gone into processing the images or selecting which ones to present. There are good images and bad images, successful experiments and failed experiments -- everything is here.
This gallery website was written using Django. I built a custom system (rather than using one of the many available image gallery packages) to better represent the interrelations between multiple data elements. Django is based on Python, which makes available NumPy, SciPy, and the Python Imaging Library, all of which are helpful in processing and displaying data.
The current system defines Datasets, Images, Data, Parameters, and Groupings. Each dataset is linked to one set of parameters, but can be associated with multiple image or data elements. Images can also link to data, and groupings can contain any number of dataset, image, and data elements.
Each experimental measurement generates either one file (for our microscopes using RHK or EasyScan software) or a group of files (for our Omicron microscope). To upload a measurement to the gallery, the file or files are processed to extract parameters and generate images. Two images are constructed automatically for each data set: a plane-subtracted image (the minimally processed image) and an image that is high-pass filtered in the fast-scan direction.
Datasets are automatically grouped into experiments, which are defined for convenience as files acquired on the same day and uploaded to the gallery at the same time; datasets are also grouped by upload batches. We also create groupings to help organize ourselves (say, to maintain an ongoing list of "good" images of molecule X on surface Y, for a particular project) or to provide a single link to communicate, "Hey, look at this" between group members.
Groupings provide a way to gather together images when working on a presentation or when preparing a manuscript. Not only does this help in organization and record-keeping, but by placing this information online, we provide (to anyone interested in finding out) the links between the original, raw experimental data and the manuscript figures in final form.
The previous incarnation of the image gallery used Zoph, an open-source project for displaying, organizing, and searching photos. Zoph did a great job of indexing and presenting our data for over 5 years, and it pretty much worked out of the box.
The code for the website is available for download here. Since this was not only my first Django project but my second Python project, the code is, frankly, a mess. You can e-mail me if you're interested in future development of this project.
The Kandel group and the University of Notre Dame retain all copyright and other rights for all images in this gallery. You may use, copy, modify, and link to any image here, under a Creative Commons Attribution License (CC-BY.)
If you present or otherwise disseminate our images in an academic or professional setting, our preferred method of attribution is:
- For informal presentations (in a classroom setting or student seminar) please use my name and affiliation, and the URL; for example, S. A. Kandel, University of Notre Dame, http://graphite.chemistry.nd.edu/view/611/1/
- For formal presentations (conference presentations or written manuscripts) please contact me so that I can supply a full author list or, when appropriate, a journal citation.
- We upload every image we acquire, good or bad, and images are either unprocessed or batch processed for easy display. Comment fields and some instrument parameters (in particular, tunneling current) are filled in at the time of acquisition (not upload) and may not be correct in all cases. I'm happy to answer questions about specific images.