Skip to content
This repository has been archived by the owner on Apr 26, 2023. It is now read-only.

Scipion demo for Buenos Aires

Jump to bottom
Laura edited this page Aug 23, 2019 · 3 revisions

Scipion demo for the course in Buenos Aires

Accessing the Virtual machines on Amazon cloud (AWS):

A number of AWS instances will be ready during the session to be used by the students. Connections details will be given at that time, and the only required software will be a web browser and a reasonable internet connection.

Get access (remember to accept the https risk):

   https://IP.xxx.yyy.zzz/vnc.html  --> pswd: scipionICS

Resize the screen with the gear at left edge menu --> remote size.

The instance type used will be a p2.xlarge, with 1 GPU Tesla K80 (12 GB RAM), 4 CPUs and 61 GB RAM.

Datasets

The dataset is the beta-galactosidase EMPIAR 10061 (4k with 38 frames).

The sample of that dataset is made of 18 movies (34 GB) and it can be found at ~/rawData.

Projects

There are two stored projects ScipionCombined and AcquisitionDemo.

ScipionCombined is a complex project that aims to show Scipion capabilities to combine different packages and protocols.

AcquisitionDemo shows a full streaming acquisition workflow already executed.

Streaming workflows

We will run a streaming workflow from a Scipion's API script.

  • All needed code is in the ~/em-facilities/usingAPI_demo directory. usingAPI_demo/scipionbox.conf contains config variables. Some of them are:

    • DEPOSITION_PATH: any directory where the script will simulate the acquisition creating symbolic links. In our case is in /microDepositions`.

    • SCIPION_PROJECT: Where to put the Scipion's project (~/ScipionUserData/projects by default).

    • RAWDATA_SIM: Where the real data is located, in this case in ~/rawData.

    • NUM_CPU: Number of CPUs of the machine. In our case 4.

The Acquisition Simulation can be launched by clicking on the AcquisitionDemo icon on the desktop (which will execute sh ~/em-facilities/usingAPI_demo/script.sh).

NOTE: To Stop it you need to close the opened terminal by the shortcut and all tasks will be killed.

Clone this wiki locally