forked from ovaexpert/ovarian-tumor-aggregation
-
Notifications
You must be signed in to change notification settings - Fork 0
/
gen-prototypes-keel.R
37 lines (32 loc) · 1.38 KB
/
gen-prototypes-keel.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
df = transmute(train.data,
I1 = train.data[, 1]+train.data[,2],
I2 = train.data[, 3]+train.data[,4],
I3 = train.data[, 5]+train.data[,6],
I4 = train.data[, 7]+train.data[,8],
I5 = train.data[, 9]+train.data[,10],
I6 = train.data[, 11]+train.data[,12],
I7 = train.data[, 13]+train.data[,14],
I8 = train.data[, 15]+train.data[,16],
I9 = train.data[, 17]+train.data[,18],
I10 = train.data[, 19]+train.data[,20],
I11 = train.data[, 21]+train.data[,22],
I12 = train.data[, 23]+train.data[,24],
Class1 = Class1,
Class2 = Class2,
Class3 = Class3,
Class4 = Class4
)
dff = bind_rows(df,
filter(df, !is.na(Class2)) %>% mutate(Class1=Class2),
filter(df, !is.na(Class3)) %>% mutate(Class1=Class3))
for(i in 1:12){
print(paste0("I",i))
tmp = filter(dff, Class1==0)[ , i]
print(paste0('Mean 0: ', round(mean(tmp[[1]])/2, 2)))
tmp = filter(dff, Class1==1)[ , i]
print(paste0('Mean 1: ', round(mean(tmp[[1]])/2, 2)))
tmp = filter(dff, Class1==2)[ , i]
print(paste0('Mean 2: ', round(mean(tmp[[1]])/2, 2)))
tmp = filter(dff, Class1==4)[ , i]
print(paste0('Mean 4: ', round(mean(tmp[[1]])/2, 2)))
}