rtidyversepropensity-score-matchingmatchit
Propensity density score with MatchIt package -- how to bind rows when we have lot of datasets to have a final dataset with matched characteristics
I'm expanding this post -- answered by @edwards (Thanks).
I'm working with panel data. We assessed children in 2019, 2020, 2021 and 2022. Therefore, I have four datasets (2019, 2020, 2021, and 2022). I want to create a fifth dataset matching data from the 2nd, 3rd, 4th datasets that matches the characteristics of the first dataset (2019).This final dataset will have fewer participants, but they'll share the same characteristics as their "peers" from 2019. The proportion of boys and girls will be about the same of 2019, the mother's age will be about the same, etc.
I read the documentation for MatchIt. I know I can work with bind_row but I'm not being able to do that.
My final comparison table should be like this
My code
data <- bind_rows(df_2019, df_2020, .id="year") |>
mutate(year=+(year==1)) # 1=2019 (treated), 0=2020 (controls)
match_obj <- matchit(year ~ asqse_quest+year_completed_cat+sex_male+momage+momed+income,
data = data,
exact= ~ momed+income,
method = "optimal")
summary(match_obj)
matched_data <- match.data(match_obj)
# Summarize the merged dataset
matched_data %>%
tableby(year_completed_cat ~ asqse_risk + momage + momed + income_poor,, data =.) %>% summary(text=T, digits=2) %>% as.data.frame()
Any suggestions are welcomed.
df_2019 = structure(list(asqse_quest = c(24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24), asqse_risk = c(0, 0, 0, 0, 0, 0, 1,
1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 0, 0, 0,
0, 1), year_completed_cat = structure(c(2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L), levels = c("18", "19", "20",
"21", "22", "23", "24"), class = "factor"), sex_male = c(1, 1,
0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1,
1, 1, 0, 1, 1, 0, 1), momage = c(35, 35, 22, 38, 33, 30, 32,
26, 21, 34, 29, 19, 35, 29, 30, 18, 30, 34, 37, 31, 23, 30, 31,
23, 27, 30, 36, 28, 28, 29), momed = c("4", "4", "2", "2", "3",
"4", "2", "2", "4", "2", "4", "2", "4", "4", "3", "2", "4", "2",
"4", "4", "3", "3", "2", "2", "2", "4", "4", "3", "2", "4"),
income = c("4", "4", "2", "3", "4", "4", "2", "3", "2", "2",
"4", "2", "4", "5", "3", "1", "4", "4", "3", "4", "3", "4",
"2", "5", "1", "4", "4", "4", "4", "5")), row.names = c(NA,
-30L), class = "data.frame", na.action = structure(c(`4` = 4L,
`14` = 14L, `18` = 18L, `22` = 22L, `40` = 40L, `46` = 46L, `51` = 51L,
`55` = 55L, `61` = 61L, `64` = 64L, `70` = 70L, `86` = 86L, `109` = 109L,
`113` = 113L, `115` = 115L, `143` = 143L, `145` = 145L, `148` = 148L,
`156` = 156L, `159` = 159L, `160` = 160L, `161` = 161L, `168` = 168L,
`175` = 175L, `205` = 205L, `206` = 206L, `209` = 209L, `221` = 221L,
`222` = 222L, `227` = 227L, `235` = 235L, `241` = 241L, `246` = 246L,
`255` = 255L, `256` = 256L, `258` = 258L, `267` = 267L, `272` = 272L,
`273` = 273L, `276` = 276L, `283` = 283L, `294` = 294L, `298` = 298L,
`304` = 304L, `311` = 311L, `312` = 312L, `316` = 316L, `341` = 341L,
`345` = 345L, `350` = 350L, `352` = 352L, `357` = 357L, `371` = 371L,
`375` = 375L, `388` = 388L, `391` = 391L, `393` = 393L, `398` = 398L,
`400` = 400L, `401` = 401L, `410` = 410L, `422` = 422L, `426` = 426L,
`441` = 441L, `446` = 446L, `450` = 450L, `456` = 456L, `459` = 459L,
`492` = 492L, `496` = 496L, `504` = 504L, `506` = 506L, `514` = 514L,
`520` = 520L, `522` = 522L, `526` = 526L, `532` = 532L, `534` = 534L,
`537` = 537L, `540` = 540L, `544` = 544L, `546` = 546L, `548` = 548L,
`557` = 557L, `562` = 562L, `565` = 565L, `575` = 575L, `577` = 577L,
`585` = 585L, `595` = 595L, `609` = 609L, `612` = 612L, `622` = 622L,
`631` = 631L, `640` = 640L, `646` = 646L, `659` = 659L, `665` = 665L,
`669` = 669L, `676` = 676L, `677` = 677L, `685` = 685L, `709` = 709L,
`710` = 710L), class = "omit"))
df_2020 = structure(list(asqse_quest = c(24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24), asqse_risk = c(0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0,
0, 0), year_completed_cat = structure(c(3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L,
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), levels = c("18", "19", "20",
"21", "22", "23", "24"), class = "factor"), sex_male = c(0, 0,
0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1,
0, 1, 1, 1, 1, 1, 1), momage = c(27, 40, 25, 35, 27, 26, 31,
31, 40, 33, 26, 21, 37, 36, 29, 30, 33, 20, 26, 33, 25, 32, 28,
34, 36, 36, 36, 39, 35, 25), momed = c("3", "4", "2", "2", "2",
"3", "4", "4", "4", "1", "2", "3", "4", "4", "2", "2", "4", "2",
"1", "4", "1", "3", "3", "4", "3", "4", "4", "3", "4", "2"),
income = c("4", "4", "4", "4", "4", "5", "4", "4", "4", "1",
"5", "3", "4", "4", "5", "4", "4", "2", "1", "4", "1", "3",
"3", "4", "5", "4", "4", "3", "5", "4")), row.names = c(NA,
-30L), class = "data.frame", na.action = structure(c(`5` = 5L,
`8` = 8L, `13` = 13L, `21` = 21L, `26` = 26L, `32` = 32L, `33` = 33L,
`34` = 34L, `42` = 42L, `46` = 46L, `51` = 51L, `57` = 57L, `58` = 58L,
`60` = 60L, `64` = 64L, `67` = 67L, `70` = 70L, `71` = 71L, `73` = 73L,
`78` = 78L, `85` = 85L, `86` = 86L, `90` = 90L, `99` = 99L, `102` = 102L,
`103` = 103L, `115` = 115L, `116` = 116L, `117` = 117L, `128` = 128L,
`129` = 129L, `131` = 131L, `139` = 139L, `147` = 147L, `151` = 151L,
`154` = 154L, `174` = 174L, `181` = 181L, `185` = 185L, `192` = 192L,
`198` = 198L, `202` = 202L, `204` = 204L, `205` = 205L, `212` = 212L,
`216` = 216L, `228` = 228L, `231` = 231L, `242` = 242L, `244` = 244L,
`250` = 250L, `252` = 252L, `253` = 253L, `257` = 257L, `282` = 282L,
`283` = 283L, `289` = 289L, `290` = 290L, `297` = 297L, `301` = 301L,
`305` = 305L, `312` = 312L, `321` = 321L, `324` = 324L, `325` = 325L,
`344` = 344L, `350` = 350L, `351` = 351L, `353` = 353L, `358` = 358L,
`363` = 363L, `364` = 364L, `374` = 374L, `375` = 375L, `379` = 379L,
`380` = 380L, `393` = 393L, `396` = 396L, `398` = 398L, `414` = 414L,
`420` = 420L, `422` = 422L, `427` = 427L, `432` = 432L, `433` = 433L,
`438` = 438L, `439` = 439L, `443` = 443L, `456` = 456L, `459` = 459L,
`461` = 461L, `465` = 465L, `468` = 468L, `471` = 471L, `482` = 482L,
`496` = 496L, `498` = 498L, `499` = 499L, `505` = 505L, `524` = 524L,
`525` = 525L, `528` = 528L, `530` = 530L, `544` = 544L, `550` = 550L,
`556` = 556L, `562` = 562L, `565` = 565L, `569` = 569L, `570` = 570L,
`571` = 571L, `578` = 578L, `579` = 579L, `601` = 601L, `613` = 613L,
`614` = 614L, `618` = 618L, `637` = 637L, `653` = 653L, `655` = 655L,
`663` = 663L, `669` = 669L, `671` = 671L, `693` = 693L, `696` = 696L,
`710` = 710L, `718` = 718L, `720` = 720L, `722` = 722L, `726` = 726L,
`731` = 731L, `733` = 733L, `736` = 736L, `744` = 744L, `761` = 761L,
`771` = 771L, `773` = 773L, `775` = 775L, `804` = 804L, `806` = 806L,
`809` = 809L, `811` = 811L, `825` = 825L, `832` = 832L, `848` = 848L,
`853` = 853L, `855` = 855L, `856` = 856L, `858` = 858L, `861` = 861L
), class = "omit"))
df_2021 = structure(list(asqse_quest = c(24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
24, 24, 24, 24, 24, 24), asqse_risk = c(0, 0, 0, 1, 0, 0, 0,
0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0,
0, 1), year_completed_cat = structure(c(4L, 4L, 4L, 4L, 4L, 4L,
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L,
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), levels = c("18", "19", "20",
"21", "22", "23", "24"), class = "factor"), sex_male = c(0, 0,
1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0,
0, 0, 0, 1, 0, 0, 0), momage = c(29, 34, 34, 25, 31, 33, 24,
26, 21, 30, 33, 41, 27, 32, 32, 30, 26, 32, 37, 20, 37, 29, 37,
31, 33, 27, 32, 33, 37, 33), momed = c("4", "2", "4", "2", "4",
"4", "3", "4", "2", "4", "2", "3", "2", "4", "4", "3", "4", "2",
"4", "2", "4", "2", "4", "2", "2", "3", "3", "4", "3", "4"),
income = c("4", "1", "4", "3", "4", "4", "2", "4", "2", "4",
"4", "4", "3", "4", "4", "4", "4", "1", "4", "3", "4", "5",
"4", "5", "4", "5", "3", "4", "4", "3")), row.names = c(NA,
-30L), class = "data.frame", na.action = structure(c(`10` = 10L,
`11` = 11L, `26` = 26L, `27` = 27L, `28` = 28L, `29` = 29L, `30` = 30L,
`31` = 31L, `32` = 32L, `33` = 33L, `42` = 42L, `53` = 53L, `66` = 66L,
`81` = 81L, `84` = 84L, `94` = 94L, `99` = 99L, `106` = 106L,
`113` = 113L, `115` = 115L, `118` = 118L, `119` = 119L, `125` = 125L,
`133` = 133L, `137` = 137L, `147` = 147L, `155` = 155L, `165` = 165L,
`170` = 170L, `171` = 171L, `188` = 188L, `189` = 189L, `192` = 192L,
`199` = 199L, `206` = 206L, `207` = 207L, `208` = 208L, `212` = 212L,
`213` = 213L, `215` = 215L, `217` = 217L, `220` = 220L, `228` = 228L,
`239` = 239L, `242` = 242L, `245` = 245L, `246` = 246L, `250` = 250L,
`255` = 255L, `256` = 256L, `267` = 267L), class = "omit"))
Created on 2024-07-13 with reprex v2.1.0
Solution
You forgot to add the data frame for year 2021.
data <- bind_rows(df_2019, df_2020, df_2021, .id="year") |>
mutate(year=+(year==1)) # 1=2019 (treated), 0=2020+2021 (controls)
Also, in your matchit call, you probably want to change the ratio argument (the default is 1:1), because you now have more data to match.
library(MatchIt)
match_obj <- matchit(year ~ asqse_quest+year_completed_cat+sex_male+momage+momed+income,
data = data,
ratio=2,
exact= ~ momed+income,
method = "optimal")
summary(match_obj)
Sample Sizes:
Control Treated
All 60. 30
Matched (ESS) 36.76 25
Matched 41. 25
Unmatched 19. 5
Discarded 0. 0
If you're still getting too few matches, and you don't care about exact matches, then omit the ones you don't care about in the call.
match_obj <- matchit(year ~ asqse_quest+year_completed_cat+sex_male+momage+momed+income,
data = data,
ratio=2,
exact= ~ momed, # + income
method = "optimal")
- Convert hex colour to HCL
- dumbell plot with more than one start and end point
- Sequence color blocks inside ggplot2 plot area at specific coordinates
- What does the symbol ::: mean in R
- Reproducing lattice dendrogram graph with ggplot2
- Make a common x-axis for multiple plots in a cowplot
- Is it possible to suppress just one tick mark label?
- R: Color branches of dendrogram while preserving the color legend
- Likert Graphs disappear when moving section labels to top instead of left?
- Adding normal density plot to histogram in ggplot
- Plotting 2 histogram on top and bottom part of a single plot
- How can I create a single paged landscape document in officedown?
- Does ctable in R not work with native pipes?
- How to find the mean number of consecutive days using dates in R?
- How to fix the value of a specific color of Legend?
- ggpplot2 join all points in a graph with a line
- Downloading Shiny Plot as PNG
- How strongly can system software interfere with R package functionality?
- R Markdown doesn't plot leaflet map in Word document
- How to switch from geom_point to geom_bin2d() or geom_tile?
- Extract non null elements from a list
- Error in passing a variable to ggplot in an r custom function
- Plotting stm model expected topic probability using ggplot2
- How to show the number of dropped observations in R?
- R: Replicating samples for a function then cumulatively adding the result
- Optional dependency in R package
- Unnest list column: In row 2, can't recycle input of size 2 to size 4
- Extract gbif occurrences within a geographic administrative area with rgbif
- PDF convert big vector file to small pixel file within R with magick
- Replace variables in a formula with their definitions