A Resource specifies a shared filesystem with one or more clusteres attached to it. This can be your local machine or just a regular cluster or even a group of cluster that can access the same FS (like Titan, Eos and Rhea do).

Once you have chosen your place to store your results t is set for the project and can (at least should) not be altered since all file references are made to match this resource.

Let us pick a local resource on your laptop or desktop machine; no cluster / HPC involved for now.

from adaptivemd import LocalResource

We now create the Resource object

resource = LocalResource()

Since this object defines the path where all files will be placed, let’s get the path to the shared folder. The one that can be accessed from all workers. On your local machine this is trivially the case.


Okay, files will be placed in $HOME/adaptivemd/. You can change this using an option when creating the Resource


Configuring your resource

Now you can add some additional paths, conda environment, etc, before we setup the project. This works by setting a special task .wrapper (see notebook 4 for more things you can do with Task objects.)

<adaptivemd.task.DummyTask at 0x110d93d50>

In a nutshell, this dummy task has a .pre and .post list of commands you can add any command you want to be executed before every task you run.

resource.wrapper.pre.append('echo "Hello World"')

A task can also automatically add to the PATH variable, set environment variables and you can add conda environments

resource.wrapper.environment['CONDA'] = 'True'
print resource.wrapper.description
Task: DummyTask
export PATH=/x/y/z:$PATH
export CONDA=True
echo "Hello World"
<main />

Let’s reset that now and just add a little comment

resource = LocalResource()
resource.wrapper.pre.append('# This is part of the adaptivemd tutorial')

Finalize the Resource

Last, we save our configured Resource and initialize our empty prohect with it. This is done once for a project and should not be altered.



LocalResource([shared_path, wrapper]) Run tasks locally and store results in $HOME/adaptivemd/