Command line¶
Once you have installed PsyNet, you interact with it by running commands in your Unix shell.
Generally speaking, you should execute all of these commands within your experiment directory
(e.g. if you are running the timeline demo: psynet/demos/timeline
).
The commands take slightly different forms depending on how you have installed PsyNet. If you are running PsyNet via Docker, you run commands that look like this:
bash docker/psynet debug local
If you are running PsyNet via virtualenv, you omit the bash docker/
and
just write commands like this:
psynet debug local
Run an experiment in debug mode (debug
)¶
The following code runs an experiment in debug mode on your local computer:
psynet debug local
The following code runs an experiment in debug mode on your own web server, via SSH;
this will push the experiment code to Heroku, but won’t recruit any participants,
even if your recruiter is set to mturk
or prolific
.
Note the specification of an app name.
psynet debug ssh --app my-app-name
This code does the same, but provisioning the web server automatically via the paid service Heroku:
psynet debug heroku --app my-app-name
Deploy an experiment (deploy
)¶
This command deploys an experiment, and enable the recruiter so you can collect real data.
psynet deploy ssh --app my-app-name # for deploying via SSH
psynet deploy heroku --app my-app-name # for deploying via Heroku
(Experimental): It is possible to deploy an experiment that resurrects the state of a previous
experiment deployment. To do this you add --archive path/to/database.zip
where
path/to/database.zip
is the path to the database.zip
file created by a previous PsyNet export.
Estimate maximum reward and completion time (estimate
)¶
This command examines the timeline, estimates how long the participant will take to complete the experiment, and how much they need to be paid as a result.
psynet estimate
Warning
This functionality is still experimental and is known to produce inaccurate results in certain cases. Always check these estimates manually before finalizing an experiment implementation.
Export data from an experiment (export
)¶
This command export data from an experiment. The data is saved by default to ~/PsyNet-data/export
.
psynet export local
psynet export ssh --app my-app-name
psynet export heroku --app my-app-name
To see further options for the export command (e.g. if you want to control the export of assets),
append --help
to these commands:
psynet export local --help
psynet export ssh --help
psynet export heroku --help
For more information on PsyNet data export see Exporting.
Generate the constraints.txt file (generate-constraints
)¶
This command generates a constraints.txt file in the experiment directory stating the exact versions of Python packages that will be installed when the server is deployed. The role of this command is still under discussion at the moment, so don’t worry too much about it.
psynet generate-constraints
Run the experiment’s regression test¶
This command runs the experiment’s regression test, as defined in test.py
. This normally involves
running one or more simulated participants through the experiment.
psynet test
Simulate data for an experiment¶
This command generates simulated data for an experiment by running the experiment’s regression test and exporting the resulting data.
psynet simulate
Update PsyNet/Dallinger (update
)¶
Note
The following command only applies if you have installed PsyNet in a local environment, rather than using Docker.
This command updates the local installations of PsyNet and Dallinger to their latest versions.
While the default is to update both packages, they can also be set to specific
versions (e.g. downgraded) using the --psynet-version
and
--dallinger-version
command line options.
psynet update
Usage
psynet update [OPTIONS]
Options:
--dallinger-version TEXT The git branch, commit or tag of the Dallinger
version to install.
--psynet-version TEXT The git branch, commit or tag of the psynet
version to install.
--verbose Verbose mode
--help Show this message and exit.