R and latex: Tabulating ranked data frame -

i generated data frame gives me top n variables , associated values melting original data frame , applying ranking function. ranked data frame looks this:

   sysid     variable      value   class1   class2 1      1  s.noun_noun 0.13121019     open     open 2      1  s.verb_verb 0.12611465     open     open 3      1    s.det_det 0.04076433   closed   closed 4      1  s.verb_noun 0.03821656     open     open 5      1   s.prep_det 0.03312102   closed   closed 6      2  s.noun_noun 0.19791667     open     open 7      2  s.verb_verb 0.13750000     open     open 8      2    s.det_det 0.04375000   closed   closed 9      2    s.prn_prn 0.03958333   closed   closed 10     2 s.coord_prep 0.03750000   closed   closed 11     3  s.noun_noun 0.16730769     open     open 12     3  s.verb_verb 0.14615385     open     open 13     3    s.det_det 0.05384615   closed   closed 14     3 s.coord_prep 0.04423077   closed   closed 15     3  s.prep_prep 0.04230769   closed   closed 

now i'm trying print latex document. i'd this:

    1                   2                    …     variable    value   variable    value    1    s.noun_noun    0.13121019  s.noun_noun 0.19791667   2    s.verb_verb    0.12611465  s.verb_verb 0.1375   3    s.det_det  0.04076433  s.det_det   0.04375  4    s.verb_noun    0.03821656  s.prn_prn   0.03958333   5    s.prep_det 0.03312102  s.coord_prep    0.0375   

i see tabular tables package potentially able this, can't figure out correct formula use. can recommend either tabular formula can format data in format, or package/function use? help.

here pretty manual way, if sysid comprised of many levels, might not seem viable option (before editing, noticed original sysid went 8). assumes data in object called df, change appropriately:

g<-split(df[,2:3], df$sysid) #here merge them side side #note variable names not right #add new row sysid ggg<- rbind.data.frame(c(g[[1]],g[[2]],g[[3]])) gggnames <- c("variable 1", "value 1", "variable 2", "value 2", "variable 3", "value 3") colnames(ggg) <- gggnames stargazer(ggg, summary=false) 

this output latex table copy , paste.

the 1 line solution library offered in comments @docendo:

 stargazer( do.call(cbind, split(df[2:3], df$sysid)), summary=false) 

again, df object of data.

if table long, xtable solution work. again, taking comments , assuming original sysid, suggests post-processing (removing bottom first table , top second table solution have in 1 table):

gggg <- do.call(cbind, split(df[2:3], df$sysid)) xtable(gggg[,1:8])   xtable(gggg[,9:16])