Skip to contents

identifying-routes

identifying-routes.Rmd
# load wide array of 
data("sio_routes_wide")
data("sio_df")

visualisation of icebeam

ranked_runs(selected_split = "ice beam", base_size = 15)
#> Error in ranked_runs(selected_split = "ice beam", base_size = 15): unused argument (base_size = 15)

visualisation of missingness

skimr::skim(sio_routes_wide %>% ungroup())
Data summary
Name sio_routes_wide %>% ungro…
Number of rows 538
Number of columns 24
_______________________
Column type frequency:
character 2
numeric 22
________________________
Group variables None

Variable type: character

skim_variable n_missing complete_rate min max empty n_unique whitespace
route_id 0 1 7 9 0 151 0
run_id 0 1 3 4 0 538 0

Variable type: numeric

skim_variable n_missing complete_rate mean sd p0 p25 p50 p75 p100 hist
morph_ball 391 0.27 198084.9 12271.25 182257 190368.0 197448.0 201975.0 267962 ▇▅▁▁▁
bombs 209 0.61 320236.4 25233.35 284423 300437.0 315811.0 330882.0 447450 ▇▆▂▁▁
charge_beam 309 0.43 581653.6 57786.52 496840 542569.0 572440.0 610169.0 884500 ▇▅▁▁▁
varia_suit 283 0.47 793834.2 105643.10 656150 703723.5 771514.0 860546.5 1223183 ▇▅▂▁▁
speed_booster 280 0.48 1043769.5 118421.34 880260 958961.8 1021397.5 1102085.8 1578300 ▇▅▂▁▁
wave_beam 287 0.47 1147976.2 136673.72 958501 1056645.5 1113824.0 1227310.0 1692016 ▇▆▂▁▁
grapple_beam 333 0.38 1406242.4 427250.80 1142252 1236463.0 1304669.0 1461615.0 3976424 ▇▁▁▁▁
ice_beam 292 0.46 2956572.0 1330944.02 941676 1596795.5 3401625.5 3938888.0 6104448 ▇▁▇▃▁
gravity_suit 220 0.59 2353624.8 344511.19 1671680 2143378.5 2282630.5 2497418.2 4366916 ▆▇▂▁▁
space_jump 164 0.70 3438227.8 551366.91 2417454 3053943.0 3350238.5 3758739.2 6571383 ▇▇▂▁▁
spring_ball 288 0.46 3735695.5 605726.21 2775254 3351968.0 3582307.5 4045408.5 6855017 ▇▆▂▁▁
plasma_beam 275 0.49 3876465.3 610386.78 2859330 3460309.5 3785841.0 4161032.0 7024767 ▇▇▂▁▁
screw_attack 182 0.66 4314722.0 704923.16 3079400 3842671.5 4186252.5 4647737.0 7967550 ▆▇▂▁▁
spazer 339 0.37 715608.2 470201.89 572126 609637.5 647279.0 709487.0 4474594 ▇▁▁▁▁
hi_jump_boots 458 0.15 951632.8 114285.33 782303 868312.0 919270.5 1009487.0 1305708 ▇▇▃▂▁
phantoon 257 0.52 1931114.8 376045.80 1298963 1680707.0 1906380.0 2071692.0 4481577 ▇▅▁▁▁
draygon 450 0.16 4021523.9 755018.09 2686882 3493542.0 3824753.0 4448553.2 6901760 ▅▇▅▁▁
ridley 400 0.26 4776457.1 967041.66 3263206 4190332.8 4637838.0 5219588.2 9442005 ▆▇▂▁▁
kraid 448 0.17 950595.4 647791.82 728782 790798.0 833321.0 951103.5 6882426 ▇▁▁▁▁
x_ray 361 0.33 5008956.8 802981.60 3956700 4406454.0 4866189.0 5401562.0 7952555 ▇▆▂▁▁
mother_brain 448 0.17 6311014.5 1169542.79 4532817 5727007.5 5956404.0 6610720.2 13574934 ▇▃▁▁▁
escape 469 0.13 5763404.3 1633863.62 97537 5498900.0 5816790.0 6502829.0 8484695 ▁▁▁▇▂

Try imputing values using KNN 

library(caret)
#> Loading required package: lattice
#> 
#> Attaching package: 'caret'
#> The following object is masked from 'package:purrr':
#> 
#>     lift

imputed_route_obs <- 
  preProcess(sio_routes_wide, 
             method="knnImpute", 
             k = 5)

imputed_route_obs$data
#> # A tibble: 0 × 22
#> # ℹ 22 variables: morph_ball <dbl>, bombs <dbl>, charge_beam <dbl>,
#> #   varia_suit <dbl>, speed_booster <dbl>, wave_beam <dbl>, grapple_beam <dbl>,
#> #   ice_beam <dbl>, gravity_suit <dbl>, space_jump <dbl>, spring_ball <dbl>,
#> #   plasma_beam <dbl>, screw_attack <dbl>, spazer <dbl>, hi_jump_boots <dbl>,
#> #   phantoon <dbl>, draygon <dbl>, ridley <dbl>, kraid <dbl>, x_ray <dbl>,
#> #   mother_brain <dbl>, escape <dbl>

How many clusters do we get over different values of k?