Size: 1690
Comment:
|
Size: 1752
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 2: | Line 2: |
2018-06-whatever: experimenting on the [[Stats test]] page |
Summary of changes so you don't have to dig through and analyze edit logs.
2018-06-whatever: experimenting on the Stats test page
2018-05-16 added ranks to the narou page and updated it
2018-05-13 fortissimo exs's stats were wrong because I was analyzing the wrong version of its script. fixed now
2018-05-13 added an example of how sampling biases frequency ranks to the frequency lists page, tl;dr: at around 30 million words the sampling is accurate to a typical deviation of +/- 2 positions at the 100th most common word, +/- 5 at the 500th, +/- 12 at the 1000th, +/- 100 at the 5000th, and so on. this is relative only to the type of work being sampled, not to the objective frequency of the word among all japanese.
2018-05-13 redumped futsuu no fantasy and added grisaia and kamikaze explorer
2018-05-12 added some new scripts to the stats (and fixed amakoi's line endings); as usual, this affects all frequency coverage and frequency target stats - some of these were hard to rip so if the stats don't make any sense just tell me
2018-04-18 updated analysis tools
2018-04-18 added amakoi
2018-04-17 added an estimate of hours taken to read based on character count, still not in the uploaded stats tools. edit: it's based on kanji/hiragana/katakana count now
2018-04-17 sjis measurement is no longer an estimate, still not in the uploaded stats tools
2018-04-16 added a sjis estimate to the stats page, not in the uploaded version of the analyzer yet
2018-04-16 added kirakira, nursery rhyme, and fortissimo exs to stats. recalculated coverages. did not update frequency list downloads yet. finally managed to statically link mecab so it runs on normal computers. updated analysis tools, minor changes.