admin管理员组文章数量:1296451
I have an absolutely enormous dataframe that I want to use most of the variables from. How can I easily make all of it go from wide to long? It seems pretty straightforward when you only really have one variable at different times, but I have quite a few variables (such as test scores in math year by year, as well as test scores in reading year by year, as well as attendance information year by year) and would like to somehow all at once make it so those are narrowed down to the ID, grade/year, and then separate columns for things like math scores, reading scores, and attendance. My goal is to do fixed effects, which isn't possible with the wide dataframe or with the narrow specifications that I've been able to elongate it to.
I want to have columns that say "math test scores" "reading test scores" etc which have the values of the results and one single column with the grade those were in, so that gk3treadss (reading score from 3rd grade) and gk2treadss (for 2nd grade) are in the same column, with separate columns denoting that student id (already in the dataset) and the grade.
I've used pivot_longer and it's worked for one column at a time, but tried to make it happen for multiple or use "unite" or "join" and it hasn't worked.
I've also used reshape but it won't separate things out properly.
There are 342 variables:
structure(list(stdntid = structure(10000, label = "STUDENT ID", format.spss = "F5.0"),
gender = structure(1, label = "STUDENT GENDER", format.spss = "F1.0", labels = c(MALE = 1,
FEMALE = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), race = structure(1, label = "STUDENT RACE/ETHNICITY", format.spss = "F1.0", labels = c(WHITE = 1,
BLACK = 2, ASIAN = 3, HISPANIC = 4, `NATIVE AMERICAN` = 5,
OTHER = 6), class = c("haven_labelled", "vctrs_vctr", "double"
)), birthmonth = structure(1, label = "STUDENT MONTH OF BIRTH", format.spss = "F2.0", labels = c(JANUARY = 1,
FEBRUARY = 2, MARCH = 3, ARPIL = 4, MAY = 5, JUNE = 6, JULY = 7,
AUGUST = 8, SEPTEMBER = 9, OCTOBER = 10, NOVEMBER = 11, DECEMBER = 12
), class = c("haven_labelled", "vctrs_vctr", "double")),
birthday = structure(22, label = "STUDENT DAY OF BIRTH", format.spss = "F2.0"),
birthyear = structure(1979, label = "STUDENT YEAR OF BIRTH", format.spss = "F4.0"),
FLAGSGK = structure(0, label = "IN STAR IN KINDERGARTEN", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG1 = structure(1, label = "IN STAR IN GRADE 1", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG2 = structure(1, label = "IN STAR IN GRADE 2", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG3 = structure(1, label = "IN STAR IN GRADE 3", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flaggk = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE K", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg1 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 1", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg2 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 2", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg3 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 3", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg4 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 4", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg5 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 5", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg6 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 6", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg7 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 7", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg8 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 8", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagprt4 = structure(1, label = "IN PARTICIPATION STUDY GRADE 4", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagidn8 = structure(0, label = "IN IDENTIFICATION STUDY GRADE 8", format.spss = "F1.0", display_width = 11L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagprt8 = structure(0, label = "IN PARTICIPATION STUDY GRADE 8", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagsatact = structure(0, label = "VALID SAT/ACT SCORE AVAILABLE", format.spss = "F1.0", display_width = 7L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flaghscourse = structure(0, label = "AT LEAST ONE YEAR OF HIGH SCHOOL COURSE DATA AVAILABLE", format.spss = "F1.0", display_width = 11L, labels = c(NO = 0,
YES = 1, MISSING = 9), class = c("haven_labelled", "vctrs_vctr",
"double")), flaghsgraduate = structure(0, label = "DATA ON HIGH SCHOOL GRADUATION STATUS AVAILABLE", format.spss = "F1.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkclasstype = structure(NA_real_, label = "CLASSROOM TYPE KINDERGARTEN", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g1classtype = structure(3, label = "CLASSROOM TYPE GRADE 1", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g2classtype = structure(3, label = "CLASSROOM TYPE GRADE 2", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g3classtype = structure(3, label = "CLASSROOM TYPE GRADE 3", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), cmpstype = structure(3, label = "CLASS TYPE COMPOSITE", format.spss = "F1.0", labels = c(SMALL = 1,
REGULAR = 2, AIDE = 3), class = c("haven_labelled", "vctrs_vctr",
"double")), cmpsdura = structure(3, label = "DURATION COMPOSITE", format.spss = "F1.0", display_width = 5L),
yearsstar = structure(3, label = "NUMBER OF YEARS IN STAR", format.spss = "F5.0", display_width = 10L),
yearssmall = structure(0, label = "NUMBER OF YEARS IN SMALL CLASSES", format.spss = "F5.0", display_width = 10L),
gkschid = structure(NA_real_, label = "KINDERGARTEN SCHOOL ID", format.spss = "F6.0"),
gksurban = structure(NA_real_, label = "SCHOOL URBANICITY KINDERGARTEN", format.spss = "F1.0", labels = c(`INNER CITY` = 1,
SUBURBAN = 2, RURAL = 3, URBAN = 4), class = c("haven_labelled",
"vctrs_vctr", "double")), gktchid = structure(NA_real_, label = "KINDERGARTEN TEACHER ID", format.spss = "F8.0"),
gktgen = structure(NA_real_, label = "TEACHER GENDER KINDERGARTEN", format.spss = "F1.0", labels = c(MALE = 1,
FEMALE = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gktrace = structure(NA_real_, label = "TEACHER RACE/ETHNICITY KINDERGARTEN", format.spss = "F1.0", labels = c(WHITE = 1,
BLACK = 2, ASIAN = 3, HISPANIC = 4, `NATIVE AMERICAN` = 5,
OTHER = 6), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkthighdegree = structure(NA_real_, label = "TEACHER HIGHEST DEGREE KINDERGARTEN", format.spss = "F1.0", labels = c(ASSOCIATES = 1,
BACHELORS = 2, MASTERS = 3, `MASTERS +` = 4, SPECIALIST = 5,
DOCTORAL = 6), class = c("haven_labelled", "vctrs_vctr",
"double")), gktcareer = structure(NA_real_, label = "TEACHER CAREER LADDER LEVEL KINDERGARTEN", format.spss = "F1.0", labels = c(`CHOSE NO TO BE ON CAREER LADDER` = 1,
APPRENTICE = 2, PROBATION = 3, `LADDER LEVEL 1` = 4, `LADDER LEVEL 2` = 5,
`LADDER LEVEL 3` = 6, PENDING = 7), class = c("haven_labelled",
"vctrs_vctr", "double")), gktyears = structure(NA_real_, label = "YEARS OF TOTAL TEACHING EXPERIENCE KINDERGARTEN", format.spss = "F2.0"),
gkclasssize = structure(NA_real_, label = "CLASS SIZE KINDERGARTEN", format.spss = "F5.0", display_width = 11L),
gkfreelunch = structure(NA_real_, label = "FREE/REDUCED LUNCH STATUS KINDERGARTEN", format.spss = "F1.0", labels = c(`FREE LUNCH` = 1,
`NON-FREE LUNCH` = 2), class = c("haven_labelled", "vctrs_vctr",
"double")), gkrepeat = structure(NA_real_, label = "REPEATING KINDERGARTEN IN 1985-1986 SCHOOL YEAR", format.spss = "F1.0", labels = c(`YES, PROMOTION RECOMMENDED` = 1,
`NO, PROMOTION NOT RECOMMENDED` = 2), class = c("haven_labelled",
"vctrs_vctr", "double")), gkspeced = structure(NA_real_, label = "SPECIAL EDUCATION STATUS KINDERGARTEN", format.spss = "F1.0", labels = c(YES = 1,
NO = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkspecin = structure(NA_real_, label = "PULLED OUT FOR SPECIAL INSTRUCTION KINDERGARTEN", format.spss = "F1.0", labels = c(YES = 1,
NO = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkpresent = structure(NA_real_, label = "DAYS PRESENT AT SCHOOL KINDERGARTEN", format.spss = "F5.0"),
gkabsent = structure(NA_real_, label = "DAYS ABSENT FROM SCHOOL KINDERGARTEN", format.spss = "F5.0"),
gktreadss = structure(NA_real_, label = "TOTAL READING SCALE SCORE SAT KINDERGARTEN", format.spss = "F5.0"),
gktmathss = structure(NA_real_, label = "TOTAL MATH SCALE SCORE SAT KINDERGARTEN", format.spss = "F5.0")), row.names = c(NA,
-1L), class = c("tbl_df", "tbl", "data.frame"))
Here is an example of the data right now:
| ID | Gender | gktreadss | g1treadss | g2treadss | gktmathss | g1tmathss | g2tmathss | gktrace | g1trace | g2trace |
|---|---|---|---|---|---|---|---|---|---|---|
| 1231 | 1 | 500 | 506 | 590 | 600 | NA | NA | 1 | 2 | 1 |
| 1232 | 2 | 533 | 533 | 690 | 655 | 667 | 700 | 2 | 1 | 1 |
| 1233 | 2 | 616 | 668 | 789 | 655 | 556 | 688 | 1 | 1 | 1 |
| 1234 | 1 | 599 | 677 | 555 | 665 | 688 | 789 | 1 | 2 | 1 |
I have an absolutely enormous dataframe that I want to use most of the variables from. How can I easily make all of it go from wide to long? It seems pretty straightforward when you only really have one variable at different times, but I have quite a few variables (such as test scores in math year by year, as well as test scores in reading year by year, as well as attendance information year by year) and would like to somehow all at once make it so those are narrowed down to the ID, grade/year, and then separate columns for things like math scores, reading scores, and attendance. My goal is to do fixed effects, which isn't possible with the wide dataframe or with the narrow specifications that I've been able to elongate it to.
I want to have columns that say "math test scores" "reading test scores" etc which have the values of the results and one single column with the grade those were in, so that gk3treadss (reading score from 3rd grade) and gk2treadss (for 2nd grade) are in the same column, with separate columns denoting that student id (already in the dataset) and the grade.
I've used pivot_longer and it's worked for one column at a time, but tried to make it happen for multiple or use "unite" or "join" and it hasn't worked.
I've also used reshape but it won't separate things out properly.
There are 342 variables:
structure(list(stdntid = structure(10000, label = "STUDENT ID", format.spss = "F5.0"),
gender = structure(1, label = "STUDENT GENDER", format.spss = "F1.0", labels = c(MALE = 1,
FEMALE = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), race = structure(1, label = "STUDENT RACE/ETHNICITY", format.spss = "F1.0", labels = c(WHITE = 1,
BLACK = 2, ASIAN = 3, HISPANIC = 4, `NATIVE AMERICAN` = 5,
OTHER = 6), class = c("haven_labelled", "vctrs_vctr", "double"
)), birthmonth = structure(1, label = "STUDENT MONTH OF BIRTH", format.spss = "F2.0", labels = c(JANUARY = 1,
FEBRUARY = 2, MARCH = 3, ARPIL = 4, MAY = 5, JUNE = 6, JULY = 7,
AUGUST = 8, SEPTEMBER = 9, OCTOBER = 10, NOVEMBER = 11, DECEMBER = 12
), class = c("haven_labelled", "vctrs_vctr", "double")),
birthday = structure(22, label = "STUDENT DAY OF BIRTH", format.spss = "F2.0"),
birthyear = structure(1979, label = "STUDENT YEAR OF BIRTH", format.spss = "F4.0"),
FLAGSGK = structure(0, label = "IN STAR IN KINDERGARTEN", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG1 = structure(1, label = "IN STAR IN GRADE 1", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG2 = structure(1, label = "IN STAR IN GRADE 2", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), FLAGSG3 = structure(1, label = "IN STAR IN GRADE 3", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flaggk = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE K", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg1 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 1", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg2 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 2", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg3 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 3", format.spss = "F8.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg4 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 4", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg5 = structure(1, label = "ACHIEVEMENT DATA AVAILABLE GRADE 5", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg6 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 6", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg7 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 7", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagg8 = structure(0, label = "ACHIEVEMENT DATA AVAILABLE GRADE 8", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagprt4 = structure(1, label = "IN PARTICIPATION STUDY GRADE 4", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagidn8 = structure(0, label = "IN IDENTIFICATION STUDY GRADE 8", format.spss = "F1.0", display_width = 11L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagprt8 = structure(0, label = "IN PARTICIPATION STUDY GRADE 8", format.spss = "F1.0", display_width = 10L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flagsatact = structure(0, label = "VALID SAT/ACT SCORE AVAILABLE", format.spss = "F1.0", display_width = 7L, labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), flaghscourse = structure(0, label = "AT LEAST ONE YEAR OF HIGH SCHOOL COURSE DATA AVAILABLE", format.spss = "F1.0", display_width = 11L, labels = c(NO = 0,
YES = 1, MISSING = 9), class = c("haven_labelled", "vctrs_vctr",
"double")), flaghsgraduate = structure(0, label = "DATA ON HIGH SCHOOL GRADUATION STATUS AVAILABLE", format.spss = "F1.0", labels = c(NO = 0,
YES = 1), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkclasstype = structure(NA_real_, label = "CLASSROOM TYPE KINDERGARTEN", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g1classtype = structure(3, label = "CLASSROOM TYPE GRADE 1", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g2classtype = structure(3, label = "CLASSROOM TYPE GRADE 2", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), g3classtype = structure(3, label = "CLASSROOM TYPE GRADE 3", format.spss = "F1.0", labels = c(`SMALL CLASS` = 1,
`REGULAR CLASS` = 2, `REGULAR + AIDE CLASS` = 3), class = c("haven_labelled",
"vctrs_vctr", "double")), cmpstype = structure(3, label = "CLASS TYPE COMPOSITE", format.spss = "F1.0", labels = c(SMALL = 1,
REGULAR = 2, AIDE = 3), class = c("haven_labelled", "vctrs_vctr",
"double")), cmpsdura = structure(3, label = "DURATION COMPOSITE", format.spss = "F1.0", display_width = 5L),
yearsstar = structure(3, label = "NUMBER OF YEARS IN STAR", format.spss = "F5.0", display_width = 10L),
yearssmall = structure(0, label = "NUMBER OF YEARS IN SMALL CLASSES", format.spss = "F5.0", display_width = 10L),
gkschid = structure(NA_real_, label = "KINDERGARTEN SCHOOL ID", format.spss = "F6.0"),
gksurban = structure(NA_real_, label = "SCHOOL URBANICITY KINDERGARTEN", format.spss = "F1.0", labels = c(`INNER CITY` = 1,
SUBURBAN = 2, RURAL = 3, URBAN = 4), class = c("haven_labelled",
"vctrs_vctr", "double")), gktchid = structure(NA_real_, label = "KINDERGARTEN TEACHER ID", format.spss = "F8.0"),
gktgen = structure(NA_real_, label = "TEACHER GENDER KINDERGARTEN", format.spss = "F1.0", labels = c(MALE = 1,
FEMALE = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gktrace = structure(NA_real_, label = "TEACHER RACE/ETHNICITY KINDERGARTEN", format.spss = "F1.0", labels = c(WHITE = 1,
BLACK = 2, ASIAN = 3, HISPANIC = 4, `NATIVE AMERICAN` = 5,
OTHER = 6), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkthighdegree = structure(NA_real_, label = "TEACHER HIGHEST DEGREE KINDERGARTEN", format.spss = "F1.0", labels = c(ASSOCIATES = 1,
BACHELORS = 2, MASTERS = 3, `MASTERS +` = 4, SPECIALIST = 5,
DOCTORAL = 6), class = c("haven_labelled", "vctrs_vctr",
"double")), gktcareer = structure(NA_real_, label = "TEACHER CAREER LADDER LEVEL KINDERGARTEN", format.spss = "F1.0", labels = c(`CHOSE NO TO BE ON CAREER LADDER` = 1,
APPRENTICE = 2, PROBATION = 3, `LADDER LEVEL 1` = 4, `LADDER LEVEL 2` = 5,
`LADDER LEVEL 3` = 6, PENDING = 7), class = c("haven_labelled",
"vctrs_vctr", "double")), gktyears = structure(NA_real_, label = "YEARS OF TOTAL TEACHING EXPERIENCE KINDERGARTEN", format.spss = "F2.0"),
gkclasssize = structure(NA_real_, label = "CLASS SIZE KINDERGARTEN", format.spss = "F5.0", display_width = 11L),
gkfreelunch = structure(NA_real_, label = "FREE/REDUCED LUNCH STATUS KINDERGARTEN", format.spss = "F1.0", labels = c(`FREE LUNCH` = 1,
`NON-FREE LUNCH` = 2), class = c("haven_labelled", "vctrs_vctr",
"double")), gkrepeat = structure(NA_real_, label = "REPEATING KINDERGARTEN IN 1985-1986 SCHOOL YEAR", format.spss = "F1.0", labels = c(`YES, PROMOTION RECOMMENDED` = 1,
`NO, PROMOTION NOT RECOMMENDED` = 2), class = c("haven_labelled",
"vctrs_vctr", "double")), gkspeced = structure(NA_real_, label = "SPECIAL EDUCATION STATUS KINDERGARTEN", format.spss = "F1.0", labels = c(YES = 1,
NO = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkspecin = structure(NA_real_, label = "PULLED OUT FOR SPECIAL INSTRUCTION KINDERGARTEN", format.spss = "F1.0", labels = c(YES = 1,
NO = 2), class = c("haven_labelled", "vctrs_vctr", "double"
)), gkpresent = structure(NA_real_, label = "DAYS PRESENT AT SCHOOL KINDERGARTEN", format.spss = "F5.0"),
gkabsent = structure(NA_real_, label = "DAYS ABSENT FROM SCHOOL KINDERGARTEN", format.spss = "F5.0"),
gktreadss = structure(NA_real_, label = "TOTAL READING SCALE SCORE SAT KINDERGARTEN", format.spss = "F5.0"),
gktmathss = structure(NA_real_, label = "TOTAL MATH SCALE SCORE SAT KINDERGARTEN", format.spss = "F5.0")), row.names = c(NA,
-1L), class = c("tbl_df", "tbl", "data.frame"))
Here is an example of the data right now:
| ID | Gender | gktreadss | g1treadss | g2treadss | gktmathss | g1tmathss | g2tmathss | gktrace | g1trace | g2trace |
|---|---|---|---|---|---|---|---|---|---|---|
| 1231 | 1 | 500 | 506 | 590 | 600 | NA | NA | 1 | 2 | 1 |
| 1232 | 2 | 533 | 533 | 690 | 655 | 667 | 700 | 2 | 1 | 1 |
| 1233 | 2 | 616 | 668 | 789 | 655 | 556 | 688 | 1 | 1 | 1 |
| 1234 | 1 | 599 | 677 | 555 | 665 | 688 | 789 | 1 | 2 | 1 |
library(tidyverse)
d <- tibble(
ID = c(1231, 1232, 1233, 1234),
Gender = c(1, 2, 2, 1),
gktreadss = c(500, 533, 616, 599),
g1treadss = c(506, 533, 668, 677),
g2treadss = c(590, 690, 789, 555),
gktmathss = c(600, 655, 655, 665),
g1tmathss = c(NA, 667, 556, 688),
g2tmathss = c(NA, 700, 688, 789),
gktrace = c(1, 2, 1, 1),
g1trace = c(2, 1, 1, 2),
g2trace = c(1, 1, 1, 1)
)
There are more years for each separate type of data and more types that I want to include.
This is what I want it to look like (obviously for more datapoints):
| ID | Gender | treadss | tmathss | trace | grade |
|---|---|---|---|---|---|
| 1231 | 1 | 500 | 600 | 1 | k |
| 1231 | 1 | 506 | NA | 2 | 1 |
| 1231 | 1 | 590 | NA | 1 | 2 |
| 1232 | 2 | 533 | 655 | 2 | k |
| 1232 | 2 | 533 | 667 | 1 | 1 |
| 1232 | 2 | 690 | 700 | 1 | 2 |
| 1233 | 2 | 616 | 655 | 1 | k |
| 1233 | 2 | 668 | 556 | 1 | 1 |
| 1233 | 2 | 789 | 688 | 1 | 2 |
| 1234 | 1 | 599 | 665 | 1 | k |
| 1234 | 1 | 677 | 688 | 2 | 1 |
| 1234 | 1 | 555 | 789 | 1 | 2 |
A major issue is just how many variables there are. Some I want to use, some I don't want to, but I'm not sure how to make sure some of them don't get messed up.
Share Improve this question edited Feb 11 at 22:43 T L asked Feb 11 at 20:27 T LT L 213 bronze badges 1 |1 Answer
Reset to default 0You can pivot_longer your simple example dataframe d like this:
df_long <- d %>%
pivot_longer(cols = matches("g(k|[12])(.*)"),
names_to = c("grade", ".value"),
names_pattern = "g(k|[12])(.*)")
giving
# A tibble: 12 × 6
ID Gender grade treadss tmathss trace
<dbl> <dbl> <chr> <dbl> <dbl> <dbl>
1 1231 1 k 500 600 1
2 1231 1 1 506 NA 2
3 1231 1 2 590 NA 1
4 1232 2 k 533 655 2
5 1232 2 1 533 667 1
6 1232 2 2 690 700 1
7 1233 2 k 616 655 1
8 1233 2 1 668 556 1
9 1233 2 2 789 688 1
10 1234 1 k 599 665 1
11 1234 1 1 677 688 2
12 1234 1 2 555 789 1
Similarly, you could pivot longer your wide data.frame df multiple times and left join it back with stdntid and grade.
library(tidyr)
library(dplyr)
library(haven)
reshape_educational_data <- function(df) {
df <- df %>%
mutate(across(where(is.labelled), ~as.numeric(as_factor(.))))
# FLAGSG type variables
flags <- df %>%
select(stdntid, matches("FLAGSG[k0-9]")) %>%
pivot_longer(
cols = matches("FLAGSG[k0-9]"),
names_to = "grade",
values_to = "flags",
names_pattern = "FLAGSG?(.*)"
) %>%
mutate(grade = ifelse(grade == "K", "0", grade))
# FLAGG type variables
flagg <- df %>% # flagg is a complicated one, it has values for 9 grades, the others don't
select(stdntid, starts_with("FLAGG")) %>%
pivot_longer(
cols = starts_with("FLAGG"),
names_to = "grade",
values_to = "flagg",
names_pattern = "flagg?(.*)"
) %>%
mutate(grade = ifelse(grade == "k", "0", grade))
# Class type variables
class_type <- df %>%
select(stdntid, matches("^g[k0-9]classtype$")) %>%
pivot_longer(
cols = matches("^g[k0-9]classtype$"),
names_to = "grade",
values_to = "class_type",
names_pattern = "g(.*?)classtype"
) %>%
mutate(grade = ifelse(grade == "k", "0", grade))
# Test scores
test_scores <- df %>%
select(stdntid, matches("^g[k0-9]t(read|math)ss$")) %>%
pivot_longer(
cols = matches("^g[k0-9]t(read|math)ss$"),
names_to = c("grade", "subject"),
values_to = "score",
names_pattern = "g(.*?)t(.*)ss"
) %>%
mutate(
grade = ifelse(grade == "k", "0", grade),
subject = case_when(
subject == "read" ~ "reading",
subject == "math" ~ "math"
)
) %>%
pivot_wider(
names_from = subject,
values_from = score,
names_glue = "{subject}_score"
)
# Attendance data
attendance <- df %>%
select(stdntid, matches("^g[k0-9](present|absent)$")) %>%
pivot_longer(
cols = matches("^g[k0-9](present|absent)$"),
names_to = c("grade", "attendance_type"),
values_to = "days",
names_pattern = "g(.*?)(present|absent)"
) %>%
mutate(
grade = ifelse(grade == "k", "0", grade)
) %>%
pivot_wider(
names_from = attendance_type,
values_from = days
)
# Teacher characteristics
teacher_vars <- df %>%
select(stdntid, matches("^g[k0-9]t(gen|race|highdegree|career|years)$")) %>%
pivot_longer(
cols = matches("^g[k0-9]t(gen|race|highdegree|career|years)$"),
names_to = c("grade", "characteristic"),
values_to = "value",
names_pattern = "g(.*?)t(.*)$"
) %>%
mutate(
grade = ifelse(grade == "k", "0", grade)
) %>%
pivot_wider(
names_from = characteristic,
values_from = value,
names_prefix = "teacher_"
)
# School characteristics
school_vars <- df %>%
select(stdntid, matches("^g[k0-9](schid|surban)$")) %>%
pivot_longer(
cols = matches("^g[k0-9](schid|surban)$"),
names_to = c("grade", "characteristic"),
values_to = "value",
names_pattern = "g(.*?)(schid|surban)"
) %>%
mutate(
grade = ifelse(grade == "k", "0", grade)
) %>%
pivot_wider(
names_from = characteristic,
values_from = value,
names_prefix = "school_"
)
# Combine all the reshaped data
long_data <- df %>%
select(all_of(c("stdntid", "gender"))) %>% # add more cols you want to keep
cross_join(
distinct(class_type, grade) %>% arrange(grade)
) %>%
left_join(class_type, by = c("stdntid", "grade")) %>%
left_join(test_scores, by = c("stdntid", "grade")) %>%
left_join(attendance, by = c("stdntid", "grade")) %>%
left_join(teacher_vars, by = c("stdntid", "grade")) %>%
left_join(school_vars, by = c("stdntid", "grade")) %>%
left_join(flags, by = c("stdntid", "grade")) %>%
arrange(stdntid, grade)
#res <- flagg %>% left_join(long_data, by = c("stdntid", "grade")) %>%
return(long_data)
}
res <- reshape_educational_data(df)
本文标签： How can I make my enormous dataframe from wide to long in RStack Overflow
版权声明：本文标题：How can I make my enormous dataframe from wide to long in R? - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1741636525a2389672.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。
df %>% pivot_longer(-c(ID, Gender), names_to = c("grade", "measure"), names_pattern = "g(k|[12])(.*)") %>% pivot_wider(names_from = measure, values_from = value) %>% relocate(grade, .after = last_col())and see if that works? – jpsmith Commented Feb 11 at 22:06