That color looks pretty nice! Would there be a way to enforce 100% as the top of the graph though?
Yo, good work so far guys, I really love to lurk here and see the progress
Now I feel bad about complaining about such a stupid thing again, but if its quick fix, i would love to actually see my longest streak as my current streak (the only streak). Currently my longest streak shows as 0 and its bugging me slightly.
I saw saraqaels screenshot of 166 current streak, 166 longest streak so i think its supposed to work like that? l guess its showing 0 for me because i didnt have single day of not doing reviews in my entire WK history so i dont really have a single finished streak to compare this ongoing streak to?
Don’t feel bad about reporting a bug, worst thing that could happen is nothing
Having 100% be the top of the chart would be even better if possible, yes!
EDIT: I also feel like the “total amount kanji percentage” literature chart would be better off having the total percentage on the y axis labels rather than the number of total kanji - “10000000” means very little to me, whereas “50%” actually gives some useful information!
I just saw all of your great feedback today and I’m gonna get to implementing it all soon. Thanks as always for your ideas and critique, as well as bug reports I’m currently on vacation but will get to it as soon as I can (likely today)!
I noticed that my daily streak is greater than my account age:
This obviously doesn’t make too much sense, so I’ve been wondering if the streak is calculated wrongly.
Either that, or this is some timezone issue (I’m UTC +1).
Anyhow, wkstats.com and the heatmap script claim that both my account age and streak length is 120 days (start date 2022-04-22)
First of all, sorry for the really long wait again I had a lot to do. But now it’s time to reveal my fixes and implementations (I hope they even work)!
I merged this pull request.
edit: Great stuff btw. Works really well!!
I added the lines to the literature charts and, as you can see the percentage as the y-axis as well!
I think that should be fixed now too. Let me know if it worked for you!
I tried to fix it but I couldn’t really reproduce the error. For me the review streak aligns with the one from the heatmap But definitely let me know how it looks now
Perfect! thanks for looking at it
dont sweat it man, its just a fun side project - its not like we are paying you to code i really hope wanikani have not completely turned from japanese learning website to coding learning website for you
It does look good now:
Thanks for taking your time to fix this
Uh-oh… don’t fall into the reorder abuse trap please!
I’m already trying to get it back on track. It got like this because I was on vacation (and using reorder script during that time) while still wanting to keep my streak alive so I didn’t activate vacation mode. Well, I guess I’ll just have to grind it down a bit before leveling up again.
Hi, this is an amazing resource - I’m getting so much information out of it.
A couple of suggestions/requests:
- In the SRS item history, I’d like to be able to ‘turn off’ the Burned items - I have 3000 burns, so all of the other levels are squashed down the bottom of the graph. In fact all I actually want to see is the Apprentice levels, perhaps if I could toggle the levels by clicking on the legend or something?
- I find the walls of fame/shame of very little value. The graphics and labels tell me very little. I think I would prefer something with more data points. Maybe top/bottom 10 items. Also I don’t get much out of the correct/incorrect ratio. I think I would prefer a measure like “Time spent at Apprentice level”.
Just some ideas - love your work
This site is getting stuck on a loading screen whenever I try open it, anyone know why or how I can fix?
Wanikani’s server has a “rate limit” for fetching data. It’s possible that you hit the rate limit. Try waiting for 1 minute and see if it resumes. (After the first time, it shouldn’t need to fetch all of the data again, so it should only be a one-time issue).
Wow this is a fantastic tool (only just stumbled across it now)! I tried to compare various text corpora with WaniKani levels as well (Anki x MorphMan if you know that) but this stylish online tool is super convenient. I love working towards the ability to read a particular novel.
Would it be silly to have a similar plot like the kanji coverage plot but for words in general? In some sense, I’d expect the number of words in a novel to be a multiple of the number of kanji but in a way it would be neat to see that on WK level x you not only have p% coverage of the kanji but also q% of the words appearing in that text.
I think this is related to the CORS policy error. For me it does not work at the moment either but I don’t exactly know why because on a different device it still works just fine. That error is notoriously elusive, I think. You could try using a different API key or a different browser and see if it works (it works in microsoft edge for me.)
A quick and dirty way if you use chrome would be to go into the search bar in windows and type “Run” and open it. Then type
chrome.exe --user-data-dir="C://Chrome dev session" --disable-web-security which would basically disable CORS. It worked for me but I can’t really say I recommend doing it because it disables the default chrome security for the tab that opened…
@rfindley Any ideas why that might not be working? The error seems to occur in the wkof where the data is fetched (“Apiv2.js”). It either gives this error below or the CORS policy error…
I also thought about that and it’d be a great addition. But I don’t really know how I could realize it yet. For the Kanji I just analyzed the raw pdf files for the books myself but it’s quite hard to filter out words from just the text of books. I also couldn’t find any premade datasets of Japanese words for specific books. But I’ll definitely try to find a way to do it if you want to have it added
I’ve never been able to reproduce it, even with the API key of people having the issue.
Ok, well then. It’ll just stay an eternal mystery. Only the CORS gods know how to fix this.
Making a new API key fixed it