remplazar valores raros

 dat <- data.frame(
  stringsAsFactors = FALSE,
  trt = c(2,2,2,2,2,2,2,10,10,10,
          10,10,10,10,2,2,2,2,2,2,2,10,10,10,10,10,
          10,10,2,2,2,2,2,2,2,10,10,10,10,10,10,10),
  par = c("2_1","2_1","2_1","2_1",
          "2_1","2_1","2_1","10_1","10_1","10_1","10_1","10_1",
          "10_1","10_1","2_2","2_2","2_2","2_2","2_2","2_2",
          "2_2","10_2","10_2","10_2","10_2","10_2","10_2",
          "10_2","2_3","2_3","2_3","2_3","2_3","2_3","2_3",
          "10_3","10_3","10_3","10_3","10_3","10_3","10_3"),
  bk = c(1,1,1,1,1,1,1,1,1,1,1,
         1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,3,3,
         3,3,3,3,3,3,3,3,3,3,3,3),
  tt = c(15,58,95,146,165,180,248,
         15,58,95,146,165,180,248,15,58,95,146,165,
         180,248,15,58,95,146,165,180,248,15,58,95,146,
         165,180,248,15,58,95,146,165,180,248),
  inc = c(1,10,20,19,16,19,19,4,
          39,55,57,49,54,54,1,13,15,51,18,15,18,1,25,
          39,44,40,43,43,1,5,13,15,13,14,15,4,40,47,
          53,10,47,45)
) 
 
dat %>% 
  ggplot()+ 
  aes(x=tt, y=inc)+ 
  geom_line()+
  geom_point()+
  facet_wrap("par") 


dat %>% 
  mutate(inc = case_when( 
    par == "10_3" & tt == "165" ~ NA_real_, 
    par == "2_2" & tt == "146" ~ NA_real_, 
    TRUE ~ inc)) %>% 
  group_by(par) %>% 
  mutate(inc = zoo::na.approx(inc, na.rm=FALSE))  -> dat2


dat2 %>% 
  ggplot()+ 
  aes(x=tt, y=inc)+ 
  geom_line()+
  geom_point()+
  facet_wrap("par")