0

I keep getting the "class and train have different lengths" error when trying to use the knn model on my dataset. 

newDF<- newDF[c(14, 1:13)]
newDF
str(newDF)

newDF1 <- newDF[c(2:11, 14)]
newDF1

df_train = newDF1[1:47385,]
dim(df_train)

df_test = newDF1[47386:59231,]
dim(df_test)


train_lbl <- newDF[1:47385,1]
test_lbl <- newDF[47386:59231,1]
dim(train_lbl)

install.packages("class")
library(class)

newDF_pred <- knn(train = df_train, test = df_test, cl = train_lbl, k = 245)

CrossTable(x = test_lbl, y=newDF_pred, propchisq=FALSE)

newDF is my entire dataset, while newDF1 is inclusive only of datatype "num"

Where is the issue and how can I fix it?

This is the data: -10lgP Mass Length ppm m/z RT start end Intensity Sample 9 Precursor Id range 1 0.543 0.234 0.245 0.348 0.0310 0.543 0.234 0.245 0.348 0.0310 0.0254 2 0.198 0.476 0.499 0.348 0.588 0.198 0.476 0.499 0.348 0.588 0.0256 3 0.234 0.245 0.348 0.0310 0.543 0.234 0.245 0.348 0.0310 0.543 0.0255 4 0.476 0.499 0.348 0.588 0.198 0.476 0.499 0.348 0.588 0.198 0.0254 5 0.245 0.348 0.0310 0.543 0.234 0.245 0.348 0.0310 0.543 0.234 0.0254 6 0.499 0.348 0.588 0.198 0.476 0.499 0.348 0.588 0.198 0.476 0.0256 7 0.348 0.0310 0.543 0.234 0.245 0.348 0.0310 0.543 0.234 0.245 0.0255 8 0.348 0.588 0.198 0.476 0.499 0.348 0.588 0.198 0.476 0.499 0.0254 9 0.0310 0.543 0.234 0.245 0.348 0.0310 0.543 0.234 0.245 0.348 0.0254 10 0.588 0.198 0.476 0.499 0.348 0.588 0.198 0.476 0.499 0.348 0.0256

... with 59,221 more rows

The size for the class and train are as follows: dim(train_lbl) [1] 47385 1

dim(df_train) [1] 47385 11

acylam
  • 18,231
  • 5
  • 36
  • 45
Heena
  • 113
  • 1
  • 2
  • 6
  • 1
    Please share a portion of your data – astrofunkswag Feb 04 '19 at 20:56
  • I added it to the bottom – Heena Feb 04 '19 at 21:49
  • The structure of the data isn't clear. See if you get the same error with just a small portion of your data, eg 20 rows, and then run `dput` which produces a nice copy-pastable version of that data. Check out this link on how to produce a [reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) if you're unsure of how – astrofunkswag Feb 05 '19 at 18:18

0 Answers0