Release 2024.5.3

Minor change to default arguments.

Release 2024.4.11

Modified the start and end timestamps for calculating resolution bins when using secondary features.

Release 2023.11.06

Changes to inactive_duration to reflect change in pandas. Has no effect on outcome

Release 2023.10.19

Fixed a minor issue with cortex.run() when a participant had no data and the start and end ranges were not specified.

Release 2023.06.26

• Have resolved an issue in screen_duration where None values were not being returned over the correct timespan

What's Changed

• updated d quality by @carlan1 in #145
• apptime updates by @carlan1 in #146
• Changes to cortex run by @carlan1 in #147
• I updated the step_count code by @lugray1 in #149
• fixing screen duraion by @carlan1 in #150
• Fixing None values screen dur by @carlan1 in #151

Full Changelog: 2023.1.9...2023.09.26

Release 2023.1.9

Created updates to call degree to match new raw sensors.

What's Changed

• call degree update by @carlan1 in #144

Full Changelog: 2022.10.24...2023.1.9

Release 2022.10.24

What's Changed

• Create CITATION.cff by @dcurrey88 in #142
• integrating raw sensorkit features by @carlan1 in #143

Full Changelog: 2022.10.11...2022.10.24

Release 2022.10.11

Updates:

• Changes to call duration and call number
• Addition of raw feature telephony to replace calls
• Addition of raw feature nearby_devices to replace bluetooth
• Minor bug fixes

2022.05.06: Visualizations and Util Code

New features and fixes:

• Correlation visualizations
• Participant-specific visualizations
• Module scheduling
• Notifications
• Database code
• Other useful functions
• Feature updates to match LAMP sensor types

Details

• We have added code to perform basic analysis to aid in the exploration of your data. Using our generalizable survey scoring code and standardized feature file formats, you can load all features of interest and:
  • Plot correlations between all features.
  • Determine if any features differ along a feature of interest (ex: Do participants that improved have higher hometime than those that did not improve in the study?).
  • Fit a simple logistic regression model to predict your feature of interest given all features.
• Participant-specific visualizations can be attached to individual participants on the data portal. These include:
  • Data coverage graphs for accelerometer and GPS.
  • Graphs of activity counts.
  • Easily understandable graphs of cortex variables broken into tertiles.
• We have added our suite of module scheduling functions that can be used to assign specific activities to the feed over time (ex: over a week) and send activity-specific messages via the chat. This suite also includes functionality to allow you to automate this process through the use of tags.
• Code for sending notifications (via the app, email, or slack) have been added to facilitate researcher to researcher and researcher to participant interaction.
• Although interacting directly with the database is not advised, there may be some cases where it is necessary. We have added a set of functions to help you with this process including code to restore deleted activities or participants, change participant parents, and map activity events to activities that have been deleted.
• Additional useful code streamlines the process of working with the API including adding sensors, updating activities, and generating lists of participant IDs.
• We have made several updates to features in Cortex:
  • New features: raw.device_motion, raw.device_state, raw.telephony
  • Updated features:
    • To deal with backward compatibility, raw.device_state will now pull both the deprecated raw.screen_state and the new raw.device_state.
    • Relatedly, primary.screen_state has been updated to work with both the new raw.device_state data and the old raw.screen_state data.
    • secondary.step_count has been updated to work with the new format of steps data. We will no longer support the old format.
  • Deprecated features: primary.sleep_periods, secondary.sleep_duration, secondary.sms_count

For more information, please reference our documentation.

2022.03.11: New features and fixes

New features and fixes

• Data quality visualizations
• Updated survey scoring
• Step count
• Healthkit sleep
• Inactive duration (a proxy for sleep)

Details:

• Data quality visualizations include activity counts, passive data quality, and features (home time, screen duration, and step counts). These graphs will be attached to the Data Portal.
• The secondary feature step_count unifies iOS and Android steps.
• Sleep information is gathered through Healthkit from watches. This data is aggregated in the secondary feature healthkit_sleep_duration. Duration of "in_bed", "in_sleep", or "in_awake" time can be computed.
• Survey scoring has been refactored to allow for grouping of questions and more advanced scoring. A scoring dictionary is passed to map questions to categories. Users can also provide mappings from responses to integers for scoring.
• inactive_duration is an early iteration of our newest sleep algorithm. It searches for the longest bouts without any spikes in accelerometer or screen activity. Inactive duration will continue to be revised in future releases.

For further information, please visit our documentation here.