circuitscape / cloudcs

Circuitscape HTML interface

Circuitscape v4

This package is now deprecated, and will no longer be maintained. Please use the latest version: Circuitscape.jl

Circuitscape is a free, open-source program which borrows algorithms from electronic circuit theory to predict patterns of movement, gene flow, and genetic differentiation among plant and animal populations in heterogeneous landscapes. Circuit theory complements least-cost path approaches because it considers effects of all possible pathways across a landscape simultaneously. We are developing Circuitscape for Mac OS X, Windows, and Linux.

Please visit the Circuitscape website for more information.

Authors

Brad McRae, Viral B. Shah, and Tanmay Mohapatra

cloudcs's Issues

Need basic user guide

We need some basic documentation that we can refer users to (doesn't need to be comprehensive, just needs to cover what's not already documented for the desktop version). One bit of feedback I got was that we need to give a step-by-step explanation of how to do batch mode.

Indent output options for current maps

Might help solve confusion about needing to click 'write current maps' if the three options after that are indented.

CloudCS fails with gz ascii grid inputs

When loading ascii grids compressed in .gz format, the following error is returned. Not sure if this is a bona fide bug since I'm not sure we currently support .gz format for CloudCS. The same files work with standalone Circuitscape, however.

Failed.
02/18/2014 06.07.22.PM:INFO:Reading maps
Storing final configuration...
Landscape size: 40000
Output maps required for - voltage: False, current: False, cumulative only: False
Parallelization requested: False
Your roles: ['user']
Verifying profile limits...
Verifying configuration...
Reading from cloud store: gdrive://200x200mag6points.asc.gz/0B2RyqHhLbsf9T1pkaE40eGJzVGs
Reading from cloud store: gdrive://200x200mag6resist.asc.gz/0B2RyqHhLbsf9UE5vZEZ3Z08wWFk
Preparing cloud store output folder: gdrive://cs/0B2RyqHhLbsf9OXhmcG1haE9wSWs

Clicking 'raster' data type twice invokes network options

If raster data type is selected and I click on the raster button, the modeling mode choices switch to the choices appropriate for network input format.

Parallelize batch mode

We should parallelize batch mode, so that each ini file runs on one processor. For now, we can just do it for multicore, but this is not too difficult to make work across lots of nodes. Especially, if resources are being paid for, it is better to spin up more machines and get the answer faster.

Sub-folders visible when browsing Google Drive

I have just one folder in my google drive, with various subfolders. When I browse for input files from cloudCS, I can see all the subfolders (including shared subfolders from other people) in the root folder. I found it confusing for awhile. Not sure if this is intentional, or even if it's changeable, but I suspect most users would prefer to see folders displayed the same way they are organized.

Current maps are included in output if run previously

If I've run a problem that creates a current map, and then do another problem that doesn't, it packages a current map in the results file.

Send notification once an offline task is completed

We could integrate with Gmail SMTP to send an e-mail notification once an offline task is completed.

Ability to cancel existing computation

Do we have the ability to cancel an existing computation and clean up all resources?

Server administration functionality

• Show logs
• Change server configuration parameters
• List/Add/Remove allowed users and their roles
• List/Terminate login sessions
• List/Terminate tasks
• Stop/Restart server

These should be available on a different section of the GUI and appear based on user role.

Offline execution

It is natural for users not to wait in front of the computer for a long running task to finish. There is also the possibility of laptops going into sleep, disconnecting the browser.

The current implementation expects users to be online. Session is cleaned up as soon as the user disconnects. I think we may need an offline execution support where a user can start a task, and log out while the task is running. The results can then be uploaded to Google Drive when done. If the user logs in before the task is complete, the running task is again re-attached to the new session and the user can see messages flowing again.

This had come to my mind earlier but had kept it aside to implement later, but looking at the execution logs I feel we may have to do this before we get others to try.

Possible to re-populate message box when monitoring background task?

When opening a background job with long calculation times, the message box will be empty until the next message appears. This can be minutes or hours depending on the job, so the user has to wait a long time to learn the status of their job. Can older messages be displayed to show the entire run history?

Nicer browser side validations / error messages

Validating the forms at the browser will give a more responsive interface and a way to display errors alongside the relevant fields.

Since validation rules would be different for cloud or local operation, it should probably be isolated to a separate script. There should be a way to do identical validations at the server, to handle misbehaving browsers.

Possible to add batch mode?

This would be a great feature for cloudCS.

Need to check for output folder

If there is no output folder selected, cloudCS begins execution and reads maps before throwing an error, which is cryptic:Error: [Errno 13] Permission denied: u'circuitscape.ini'

It would be better to check for a valid output path before beginning execution.

CNAME for compute.circuitscape.org

It would be great if a user can point to http://compute.circuitscape.org/ and have it forward to wherever the cloudCS happens to be hosted. @bmcrae should have the DNS registration.

@tanmaykm What all configuration changes would be needed to make this happen?

Changes to wording in tabs

Some minor wording changes- if I can just make changes in templates\cs.html I am happy to do so.

Tab 1:
Under 'Choose a source/ground modeling mode' it would be good to indent parwise, one-to-all, etc slightly.

Tab 2:
Change tab name from 'Habitat Data' to 'Resistance Data'
Change 'Habitat Map' to 'Resistance File'
Change 'Mask Habitat With' to 'Mask Resistance Map with'
Change 'as per cell' to 'values coded as'
Disable mask file if in network mode

Tab 3:
Change 'Current Sources' to 'Current Source File'
Change 'Ground Points' to 'Ground Point File'
Change 'with values as' to 'values coded as'

Tab 4:
Change 'Output Options' to 'Output File Path'
Change 'Write cumulative current map only' to ' Write cumulative current map (and maximum if selected) only.
-Move above just below 'Write maximum of current maps' option