This & That

# Correlations 2.0 ## What's New ### Medications Support With iOS 26 it's now possible to access medications data and build correlations with other data types. See correlations and causal relationships between medication doses and other health data types. ### New Liquid Glass Design New Liquid Glass design for iOS 26 ### Enhanced Correlation Analysis New 'Causal only' filter and causal labels in correlation lists help you identify potential cause-and-effect relationships ### Additional Health Data Types Added support for State of Mind and tooth brushing data

- Added Workout data types, one per activity. For example "Walking Workouts", "Downhill Skiing Workouts", etc.

- Optional Arc Timeline app location data import, for historical weather data - Note: This update will do a full resync from Health app, to clean up some date+timezone storage mistakes in a previous update

- Scatter plot points are now coloured by age

- Found a way to *massively* speed up the correlations updating! Yay! - Fixed weather type to weather type correlations not being queued for update

## Weather Data Types This update adds weather data types! This & That will ask for permission to receive approximate location updates, and if granted will then start collecting daily weather reports, grouped into data types such as "Daytime Humidity", "Nighttime Temperature", etc. Note that correlations require at least three days of data to start calculating. And typically it requires several months of data before correlations start showing sensible results instead of nonsense, so it'll be a while before we see the fruits of this labour. But it'll be good to get that ball rolling! Note that the gathered location data is only approximate, using iOS's "imprecise location" system, which is designed to make it impossible to determine your exact location, for greater privacy. Additionally, all location data is deleted after being used for fetching weather reports, so that the app doesn't retain any location data longer than necessary. ## Other Changes - Made more improvements to the correlation updating, so that you should see the data you're currently looking at be updated more quickly - Restyled the lag filters bar on Correlations List view and moved it to the bottom

This update adds Granger causality calculation! Check the new analysis section at the bottom of the Correlation view. Tap the help button next to the analysis to get a simple explanation of what Granger causality means and how it can be useful.

## Added proper two column layout to the iPad app This is just the first small step into building a really fun iPad app! The iPad's larger screen area opens up possibilities for a bunch of larger, more complex, or interactive visualisations that don't really fit on the smaller iPhone screen. I wanted to first get this release live with the two column layout, even though it's only a small first step, because to understand what will work well on iPad first needs us using the app on iPad. As always, I'm happy to brainstorm ideas on the support forum! ## Fixed a bug with Total Sleep collation for days with multiple data sources Note that this fix will only impact new data. To purge and recalculate any previously incorrect Total Sleep values you will need to delete and reinstall the app, to get a full fresh HealthKit sync.

Note: This update will delete your database and do a fresh import. The database schema changed quite a bit, such that a migration wasn't really feasible. You'll need to mark your favourite data types as favourites again. Sorry about that! - Added new iOS 17 Health data types - Added lag filter buttons to Correlation List view - Fixed day values getting lost on timezone change - Fixed sleep session data not being imported if missing a timezone - Misc fixes and improvements Addendum: I would love to spend more time on developing this app! As of yet the app isn't earning enough money to justify spending a whole lot of time on it, but hopefully that will change over time. There's lots of cool things that could be done with the available data, both in visualisations and further analysis. For anyone with thoughts or ideas or questions, please do head to the support forum (https://support.bigpaua.com), where I'm always available to discuss and plan exciting ideas!

Fixed some correlations not getting marked for update.

- Massively sped up correlation updating - Fixed bug that caused some correlations to never be updated - Correlation List view's chosen lags are now based on "best" / most credible rather than simply strongest - CCF / Lagged Correlations plot's markers for strongest positive / negative lags also now use the "best" / most credible calculation instead of simply strength - Made more progress on stopping low n correlations from incorrectly floating to the top of the list

## Data Types view - The search bar now says visible when list is scrolled down from top - Now keeping the “updated” badge for 15 minutes - 5 minutes was too short ## Correlations List view - Added an "Updating correlations" indicator - Added a blue dot to correlation rows to indicate recently updated ## Correlation view - Increased calculated lags to -/+ 60 days - Added a drag gesture to the "Lagged correlations" plot to make it easier to select which lag to view - The "Lagged correlations" plot now shows a line to indicate the currently updating lag

- Added alcohol consumption data type - Added the missing lab/test result data types - Increased the calculated lags to -9 to 9 - Added a Lagged Correlations Plot (CCF plot) to Correlation View - Default to showing strongest lagged correlation on Correlation View - Improved the Top Correlations mini lists on the main Data Types view - Prioritise the correlations update queue by doing currently visible data types first