[LCWO LOGO]  

Login

User name:
Password:


Language
Български Português brasileiro
Bosanski Català
繁體中文 Česky
Dansk Deutsch
English Español
Suomi Français
Ελληνικά Hrvatski
Magyar Italiano
日本語 한국어
Bahasa Melayu Nederlands
Norsk Polski
Português Română
Русский සිංහල
Slovenščina Srpski
Svenska ภาษาไทย
Türkçe Українська
简体中文
Who is online? (20)


LCWO Discussion Forum [Atom LCWO Forum Feed]

This is a simple discussion forum for LCWO users. Feel free to use it for any kind of discussion related to this website.

Thread: Levenshtein-Errors are (grossly) wrong

Back to the Forum

AuthorText


Posted: 2021-04-25 09:33
test case: I selected lesson 6 at 20/15 WPM. Didn't even input a single word in the box and pressed "Check results".

After the warning "The number of groups sent was different from the text which you entered", I pressed the button again and I got:



Groups: 82 (385 characters), Errors 385 = 100%

Real speed: 385 characters / 309.5 seconds = 14.9 WPM / 74.6 cpm

Levenshtein-Errors: 255 = 66.2 %



How come I only have 255 errors out of 385 characters? This should be 100% errors, not 66.2%

I think the Levenshtein measurement grossly underestimates the number of errors we make.


Posted: 2021-04-25 13:37
Also in this other test, the difference between standard errors and Levenshtein-Errors is massive:

Groups: 76 (322 characters), Errors 41 = 12.7%

Real speed: 322 characters / 277.5 seconds = 13.9 WPM / 69.6 cpm

Levenshtein-Errors: 19 = 5.9 %



This is the full transcript of the lesson:



Result (20/14 WPM):
Sent Group Received Group Errors

KUMMRNS KUMMRNS KUMMRNS 0
NUN NUN NUN 0
MN MN MN 0
MESS MESS MESS 0
RKSNEE RKSNEE RKSNEE 0
KR KR KR 0
SEKMR SEKMR SEKMR 0
EMUEE EMUEE EMUEE 0
SNSSM ENSSM SNSSM 1
KERENSM KERENSM KERENSM 0
ERKNMES ERNKMES ERKNMES 2
NU NU NU 0
UMS UMS UMS 0
KKRESU KKRESU KKRESU 0
EN EN EN 0
UENEN UNSEN UENEN 2
ESMKSU SUMKSU ESMKSU 2
MU MU MU 0
KUSKN KUSKM KUSKN 1
UKNUNE UKNUNE UKNUNE 0
ME ME ME 0
RKK RKK RKK 0
RSU RSU RSU 0
MEESKS MEENSU MEESKS 3
SSU SSU SSU 0
EKE EKE EKE 0
NKURRS NSKURRS N-KURRS 1
KKUSE KKUSE KKUSE 0
KKS KKU KKS 1
MMKS MMKS MMKS 0
SNK SNK SNK 0
RMMNU RMMKU RMMNU 1
KK KK KK 0
MNNMM MKNMM MNNMM 1
EE EE EE 0
SUUUM SUSSM SUUUM 2
SSESS SSESS SSESS 0
NE NE NE 0
KRESEMM KRESEMM KRESEMM 0
EEUS EEUS EEUS 0
SEMSS SEMSS SEMSS 0
UNKNK USNK UNKNK 2
UNMRK UNMRK UNMRK 0
KRMNS KRMNS KRMNS 0
SNRUMN SNRUMN SNRUMN 0
NMU NMU NMU 0
MSSRMK MSSRMK MSSRMK 0
EKRSKSE EKRSKSE EKRSKSE 0
ESUMSUS ESUMSUS ESUMSUS 0
NURSR NURSR NURSR 0
NKM NKM NKM 0
MK MK MK 0
KSK SUM KSK 3
KREN KREN KREN 0
UKSKRMU UKSKRNU UKSKRMU 1
NEUUE NEUUK NEUUE 1
SSER SSUR SSER 1
NKS KMR NKS 3
KSKSR KSKRT KSKSR 2
NMN NMK NMN 1
KKS KKD KKS 1
MSR MSR MSR 0
SREE SRNN SREE 2
RSUU RUU RSUU 1
NU KU NU 1
KNS KNS KNS 0
RSNSNU RSNSNU RSNSNU 0
RKRM RKRM RKRM 0
RKRRK RKRRK RKRRK 0
EUNSE EUNSE EUNSE 0
EEEN EEEN EEEN 0
RSS REE RSS 2
ENKUER ENKUER ENKUER 0
MKU MKU MKU 0
RS RU RS 1
NNER MMER NNER 2
Administrator


Posted: 2021-04-30 19:50
LS errors are only calculated over the first 255 characters. This is a limitation of the levenshtein implementation of PHP.

Instead of showing a wrong error count, I will now suppress it completely for texts that exceed 255 characters (this has been the case for the "code groups" part of the site for a long time already, and I somehow forgot to do the same for the lessons).

TNX for the heads-up!

73
Fabian

Back to the Forum

You must be logged in to post a message.