* (FSRS) Improve presentation of log loss and RMSE
The number of decimal places in log loss have been increased from 3 to 4. Can help in better comparison of weights especially when the log loss with both are same up to 3 places of decimal. Also makes it consistent with the Python optimizer.
RMSE has been expressed in percent, making it easier to interpret.
* Fix formatting
---------
Co-authored-by: Damien Elmes <gpg@ankiweb.net>
* Draft set optimal/calculated retention button
Temporarily save the calculated optimal retention and display it with a button
that sets the desired retention above to this value.Don't show button until
attention had been calculated. Disable button when optimal and desired
attention are equal.
I find this nicer than the current alert-popup solution, as it avoids a popup
and gives a choice to the user to accept the calculated retention or not, while
also persisting the calculated retention on the screen for a bit.
TODO: What's still missing is that the `optimalRetention` variable is global and
persists when I change presets. When changing presets the variable should reset to
`undefined`, which would also makes the button disappear. Ideally it should also
disappear when changing the FSRS parameters. So probably it should be made part
of some deck options state and subscribe to some events. But with that I might
need some help. Also I thought whether that variable should go into the deck
options schema but tbh it's not something we want to persist between sessions,
users should recalculate it.
* Add me to contributors for tests so pass
* Add formatting ant type fixes to make tests pass
* Minor fixes (dae)
* Remove the period
It's very helpful having a sub-permille precision in a progress
indicator, percent-precision or at most a tenth of a percent should be
sufficient for any indicator.
But in particular the compute-retention progress has 10 steps, i.e. the progress
increases in 10%-intervals (10%, 20%, ...), it *cannot* have sub-decimal
progress-percentages, see 2d5b19b494/src/optimal_retention.rs (L365-L368). So there integer percents should be
enough, everything else is misleading.
The compute-weights progress is currently (as of beta-2) not showing up at all.
Maybe if the bug is fixed it can show sub-percent percentages, so for know I changed that
to 0.1% precision. But I think integer percentages should be fine here as well,
so upon request I can fix that.
Also see my comment on this problem in https://forums.ankiweb.net/t/anki-23-10-beta/34912/39.
* Pack FSRS data into card.data
* Update FSRS card data when preset or weights change
+ Show FSRS stats in card stats
* Show a warning when there's a limited review history
* Add some translations; tweak UI
* Fix default requested retention
* Add browser columns, fix calculation of R
* Property searches
eg prop:d>0.1
* Integrate FSRS into reviewer
* Warn about long learning steps
* Hide minimum interval when FSRS is on
* Don't apply interval multiplier to FSRS intervals
* Expose memory state to Python
* Don't set memory state on new cards
* Port Jarret's new tests; add some helpers to make tests more compact
https://github.com/open-spaced-repetition/fsrs-rs/pull/64
* Fix learning cards not being given memory state
* Require update to v3 scheduler
* Don't exclude single learning step when calculating memory state
* Use relearning step when learning steps unavailable
* Update docstring
* fix single_card_revlog_to_items (#2656)
* not need check the review_kind for unique_dates
* add email address to CONTRIBUTORS
* fix last first learn & keep early review
* cargo fmt
* cargo clippy --fix
* Add Jarrett to about screen
* Fix fsrs_memory_state being initialized to default in get_card()
* Set initial memory state on graduate
* Update to latest FSRS
* Fix experiment.log being empty
* Fix broken colpkg imports
Introduced by "Update FSRS card data when preset or weights change"
* Update memory state during (re)learning; use FSRS for graduating intervals
* Reset memory state when cards are manually rescheduled as new
* Add difficulty graph; hide eases when FSRS enabled
* Add retrievability graph
* Derive memory_state from revlog when it's missing and shouldn't be
---------
Co-authored-by: Jarrett Ye <jarrett.ye@outlook.com>
* Remember original id when importing notetype
* Reuse notetypes with matching original id
* Add field and template ids
* Enable merging imported notetypes
* Fix test
Note should be updated if the incoming note's notetype is
remapped to the existing note's notetype.
On the other hand, it should be skipped if its notetype id is mapped
to some new notetype.
* Change field and template ids to i32
* Add merge notetypes flag to proto message
* Add dialog for apkg import
* Move HelpModal into components
* Generalize import dialog
* Move SettingTitle into components
* Add help modal to ImportAnkiPackagePage
* Move SwitchRow into components
* Fix backend method import
* Make testable in browser
* Fix broken modal
* Wrap in container and fix margins
* Update commented Anki version of new proto fields
* Check ids when comparing notetype schemas
* Add tooltip for merging notetypes.
* Allow updating notes regardless of mtime
* Gitignore yarn-error.log
* Allow updating notetypes regardless of mtime
* Fix apkg help carousel
* Use i64s for template and field ids
* Add option to omit importing scheduling info
* Restore last settings in apkg import dialog
* Display error when getting metadata in webview
* Update manual links for apkg importing
* Apply suggestions from code review
Co-authored-by: Damien Elmes <dae@users.noreply.github.com>
* Omit schduling -> Import all cards as new cards
* Tweak importing-update-notes-help
* UpdateCondition → ImportAnkiPackageUpdateCondition
* Load keyboard.ftl
* Skip updating dupes in 'update alwyas' case
* Explain more when merging notetypes is required
* "omit scheduling" → "with scheduling"
* Skip updating notetype dupes if 'update always'
* Merge duplicated notetypes from previous imports
* Fix rebase aftermath
* Fix panic when merging
* Clarify 'update notetypes' help
* Mention 'merge notetypes' in the log
* Add a test which covers the previously panicking path
* Use nested ftl messages to ensure consistency
* Make order of merged fields deterministic
* Rewrite test to trigger panic
* Update version comment on new fields
* Support searching for deck configs by name
* Integrate FSRS optimizer into Anki
* Hack in a rough implementation of evaluate_weights()
* Interrupt calculation if user closes dialog
* Fix interrupted error check
* log_loss/rmse
* Update to latest fsrs commit; add progress info to weight evaluation
* Fix progress not appearing when pretrain takes a while
* Update to latest commit
