Foraging behavior browser

A streamlit app for browsing foraging behavior sessions in AIND.

The app

• The one running on ECS: https://foraging-behavior-browser.allenneuraldynamics-test.org/
  • If you see a window "Select a certificate for authentication", just click "Cancel".
• To debug, use this Code Ocean capsule
• See also a streamlit app for ephys sessions https://foraging-ephys-browser.allenneuraldynamics-test.org/ (Git repo)

Sharing a contextual app

Starting from this PR, the URL contains (part of) the session state of the app. Meaning that the user can "query" pre-set filters and plot settings in the URL. On the other hand, after interacting with the app, the URL is automatically updated to reflect the user interactions, and the user can then copy and paste the URL to share/save the new context. Essentially, this becomes a cool way of sharing a data analysis.

For example, this URL show all plots of mouse 699982

https://foraging-behavior-browser.allenneuraldynamics-test.org/?filter_subject_id=699982&session_plot_mode=all+sessions+filtered+from+sidebar&tab_id=tab_session_inspector

and this URL will show exactly the plot below. Note the filters and plot settings are preserved.

https://foraging-behavior-browser.allenneuraldynamics-test.org/Old_mice?filter_h2o=HH&filter_session=1.0&filter_session=81.0&filter_finished_trials=825.6&filter_finished_trials=1872.0&filter_foraging_eff=0.793295&filter_foraging_eff=1.2966&filter_task=coupled_block_baiting&filter_photostim_location=None&tab_id=tab_session_x_y&x_y_plot_xname=foraging_eff&x_y_plot_yname=finished_trials&x_y_plot_group_by=h2o&x_y_plot_if_show_dots=True&x_y_plot_if_aggr_each_group=True&x_y_plot_aggr_method_group=linear+fit&x_y_plot_if_aggr_all=False&x_y_plot_aggr_method_all=mean+%2B%2F-+sem&x_y_plot_smooth_factor=5&x_y_plot_if_use_x_quantile_group=False&x_y_plot_q_quantiles_group=20&x_y_plot_if_use_x_quantile_all=False&x_y_plot_q_quantiles_all=20&x_y_plot_dot_size=21&x_y_plot_dot_opacity=0.4&x_y_plot_line_width=3.5&auto_training_history_x_axis=session&auto_training_history_sort_by=progress_to_graduated&auto_training_history_sort_order=descending

So far, theses are all supported fields in the URL query:

foraging-behavior-browser/code/Home.py

Lines 54 to 85 in 3f4124e

	to_sync_with_url_query = {
	'filter_subject_id': '',
	'filter_session': [0.0, None],
	'filter_finished_trials': [0.0, None],
	'filter_foraging_eff': [0.0, None],
	'filter_task': ['all'],
	'tab_id': 'tab_session_x_y',
	'x_y_plot_xname': 'session',
	'x_y_plot_yname': 'foraging_eff',
	'x_y_plot_group_by': 'h2o',
	'x_y_plot_if_show_dots': True,
	'x_y_plot_if_aggr_each_group': True,
	'x_y_plot_aggr_method_group': 'lowess',
	'x_y_plot_if_aggr_all': True,
	'x_y_plot_aggr_method_all': 'mean +/- sem',
	'x_y_plot_smooth_factor': 5,
	'x_y_plot_if_use_x_quantile_group': False,
	'x_y_plot_q_quantiles_group': 20,
	'x_y_plot_if_use_x_quantile_all': False,
	'x_y_plot_q_quantiles_all': 20,
	'x_y_plot_dot_size': 10,
	'x_y_plot_dot_opacity': 0.5,
	'x_y_plot_line_width': 2.0,
	'auto_training_history_x_axis': 'date',
	'auto_training_history_sort_by': 'subject_id',
	'auto_training_history_sort_order': 'descending',
	'auto_training_curriculum_name': '',
	'auto_training_curriculum_version': '',
	'auto_training_curriculum_schema_version': '',
	}

Develop in Code Ocean

1. Duplicate the capsule foraging-behavior-browser

2. Start a VS Code machine

3. Click "Start Debugging" or press F5

   

4. You should see something like this in the terminal

   

   and a dialog like this

   

   Press "Open in Browser" will initiate the app.

5. If "Open in Browser" doesn't show up, click the browser icon in Ports - 8501 - Local Address.

   

6. You can start to debug the app by adding break points etc.