Quick way to run R-squared analysis in R

[hopnstop]

I want to calculate the R-Squared of each of the stats in columns 3-5 (targets, receptions, and yards) and how it correlates to following-year fpts. Below is a sample of the type of data I'm working with:

player	year	targets	receptions	yards	fpts
Michael Gallup	2017	113	66	1107	147
D.J. Moore	2017	135	87	1175	144
Robert Woods	2017	139	90	1134	143
Calvin Ridley	2017	93	63	866	134
Terry McLaurin	2017	93	58	919	134
Tyler Boyd	2017	148	90	1046	133
Deebo Samuel	2017	81	57	802	132
Michael Gallup	2018	91	62	779	132
D.J. Moore	2018	89	58	860	130
Robert Woods	2018	127	83	997	130
Calvin Ridley	2018	100	58	900	129
Terry McLaurin	2018	133	74	1035	127
Tyler Boyd	2018	97	66	869	124
Deebo Samuel	2018	84	48	740	122
Michael Gallup	2019	122	78	833	120
D.J. Moore	2019	106	67	778	118
Robert Woods	2019	105	54	627	118
Calvin Ridley	2019	90	49	1001	112
Terry McLaurin	2019	96	52	779	110
Tyler Boyd	2019	90	47	775	110
Deebo Samuel	2019	92	59	680	104

I'd like to get it so it appears like this, with the "next_year_fpts" column

player	year	targets	receptions	yards	fpts	next_year_fpts
Michael Gallup	2017	113	66	1107	147	132
D.J. Moore	2017	135	87	1175	144	130
Robert Woods	2017	139	90	1134	143	130
Calvin Ridley	2017	93	63	866	134	129
Terry McLaurin	2017	93	58	919	134	127
Tyler Boyd	2017	148	90	1046	133	124
Deebo Samuel	2017	81	57	802	132	122
Michael Gallup	2018	91	62	779	132	120
D.J. Moore	2018	89	58	860	130	118
Robert Woods	2018	127	83	997	130	118
Calvin Ridley	2018	100	58	900	129	112
Terry McLaurin	2018	133	74	1035	127	110
Tyler Boyd	2018	97	66	869	124	110
Deebo Samuel	2018	84	48	740	122	104
Michael Gallup	2019	122	78	833	120	NA
D.J. Moore	2019	106	67	778	118	NA
Robert Woods	2019	105	54	627	118	NA
Calvin Ridley	2019	90	49	1001	112	NA
Terry McLaurin	2019	96	52	779	110	NA
Tyler Boyd	2019	90	47	775	110	NA
Deebo Samuel	2019	92	59	680	104	NA

This way I'll be able to calculate R-Squared for the targets, receptions, etc. in year x compared to the fpts in year x+1. I'll have several years of data with many more players, so looking for quick and reproducible way to make my data the way that I need it.

[aghaynes]

You could copy your dataframe, replace year by year - 1 and then merge the columns back in...

dat2 <- dat
dat2$year <- dat2$year-1
dat2$next_year_fpts <- dat2$fpts
dat <- merge(dat, dat2[, c("player", "year", "next_year_fpts")])

not a tidyverse solution, but it should work, and there would be an analogous solution with tidyverse I'm sure.

[HanOostdijk]

Tidyverse solution:

df = tibble::tibble(
  player = c('a','b', 'a','b', 'a','c'),
  year   = c(2017, 2017, 2018, 2018, 2019,2019),
  targets= c(113,114,115,116,115,100),
  fpts   = c(147,148,132,144,130,12)
)
df
#> # A tibble: 6 x 4
#>   player  year targets  fpts
#>   <chr>  <dbl>   <dbl> <dbl>
#> 1 a       2017     113   147
#> 2 b       2017     114   148
#> 3 a       2018     115   132
#> 4 b       2018     116   144
#> 5 a       2019     115   130
#> 6 c       2019     100    12

tidyr::pivot_wider(df,id_cols=player,names_from=year,values_from=c(fpts,targets))
#> # A tibble: 3 x 7
#>   player fpts_2017 fpts_2018 fpts_2019 targets_2017 targets_2018 targets_2019
#>   <chr>      <dbl>     <dbl>     <dbl>        <dbl>        <dbl>        <dbl>
#> 1 a            147       132       130          113          115          115
#> 2 b            148       144        NA          114          116           NA
#> 3 c             NA        NA        12           NA           NA          100

^{Created on 2020-06-04 by the reprex package (v0.3.0)}