R/pretty_mixed_models.R
In z2thet/MoffittFunctions: Moffitt Functions
Defines functions
pretty_mixed_effects_coxme
pretty_mixed_effects_glmer
Documented in
pretty_mixed_effects_coxme
pretty_mixed_effects_glmer
#' Fancy Table Output of Mixed effects model from glmer
#'
#' pretty_mixed_effects_glmer() takes a mixed effects model from glmer() fit object and calculates estimates, odds ratios with confidence intervals. P values are also produced. For categorical variables with 3+ levels overall Type 3 p values are calculated, in addition to p values comparing to the first level (reference).
#'
#' @param fit glmer fit
#' @param df is data.frame or tibble used to create model fits. Used for capturing variable labels, if they exist
#' @param rdintterm random term in the model
#'
#' @return
#'
#' A data.frame with: \code{Variable}, \code{Level}, \code{ OR (95\% CI)}, \code{P Value} (for categorical variables comparing to reference), \code{Overall P Value} (for categorical variables with 3+ levels).
#'
#' @importFrom dplyr %>%
#' @importFrom dplyr case_when mutate group_by select ungroup
#' @importFrom stats anova
#' @importFrom broom.mixed tidy
#'
#' @examples
#'
#'# glmer fit ####
#' library(magrittr)
#' library(ggeffects)
#' library(sjmisc)
#' library(lme4)
#' library(splines)
#' library(coxme)
#' library(car)
#' library(survival)
#' library(broom.mixed)
#' library(rlang)
#' library(kableExtra)
#' data(PDACdata)
#' gmlmerfit <- glmer(I(PDACdata$mortality90day)=="Dead" ~ caci3
#' + surgery_volume_recode
#' + (1|puf_facility_id)
#' ,family = binomial(link = "logit"),
#' data = PDACdata )
#'
#' # CALL function include fit name, data frame name and random intercept term name in quotes ####
#' outputtable <- pretty_mixed_effects_glmer(gmlmerfit, PDACdata, "puf_facility_id")
#'
#' kableExtra::kable(outputtable, 'html', caption = 'My Table') %>%
#' kableExtra::collapse_rows(c(1:2), row_group_label_position = 'stack')
#'
#' # PDF output
#' #kable(outputtable, 'latex', escape = FALSE, longtable = T, booktabs = TRUE,
#' # linesep = '', caption = 'Multivariable Mixed Effects Logistic Regression
#' # Model Results for 90 day mortality with CACI, Volume, and Facility type')%>%
#' # footnote(number = c('OR are odds of death within 90 days
#' # after the most definitive primary site surgery'))
#'
#'
#' @export
pretty_mixed_effects_glmer <- function(fit,df,rdintterm){# fit is glmer object, and df = data frame is the data used.
#rdintterm = is the name of the variable used for random intercept in quotes
df<-df;
fitresults<-fit;
confidenceintervalsexp2<-tidy(fit,conf.int=TRUE, exponentiate=TRUE, effects="fixed");
car::Anova(fit, test = "Chisq") -> LRfitanova;
LRanovastats <- as.data.frame(cbind(Variable = rownames(LRfitanova),
'Overall P' = MoffittFunctions::pretty_pvalues(LRfitanova$'Pr(>Chisq)') ));
reppedvec <- function(fitresults,df){
facid<-match("puf_facility_id",names(attributes(fit)$frame));
thevars <-names(attributes(fit)$frame)[c(-1,-facid)]#attributes(LRmodel_rint0$terms)$term.labels;
thevars <- gsub("`","", thevars)
nvars <- length(thevars);
therepeats <- rep(NA,nvars)
for (i in 1:nvars){ therepeats[i] <- ifelse("numeric" %in% class(df[,thevars[i]]), 1, length(unique(df[,thevars[i]]))- 1) }
return(therepeats)
}
# NEED match names to get rid of facilty id ####
facid<- match(rdintterm, names(attributes(fit)$frame))
Variable <- rep(names(attributes(fit)$frame)[c(-1,-facid)], times = reppedvec(fit,df ) )
confidenceintervalsexp2$Variable <- c("Intercept",Variable);
lrmetable <- as.data.frame( dplyr::left_join(confidenceintervalsexp2, LRanovastats, by = "Variable") %>%
mutate( 'P Value' = MoffittFunctions::pretty_pvalues(p.value)) %>%
group_by(Variable) %>%
mutate(first = 1:dplyr::n(),
total = dplyr::n()) %>%
mutate(
'Overall P' = case_when(first == 1 & (first != total) ~ as.character(`Overall P`) ,
first == total ~ "",
first != total ~ "" ))%>%
ungroup() %>%
mutate(Variable = as.character(Variable),
Level = as.character(term) )
)#end as data .frame
# remove variable names from levels ####
test <- rep(NA,NROW(lrmetable))
for(i in 1:NROW(lrmetable)){
test[i] <- ifelse((lrmetable$first[i] == 1 & (lrmetable$first[i] == lrmetable$total[i])) ,
" ",stringr::str_replace(lrmetable[i,"Level"], lrmetable[i,"Variable"], "") ) }
lrmetable$test <- test
lrmetable <- lrmetable %>%
mutate( Level = test) %>%
mutate( Variable = case_when(first == 1 ~ as.character(Variable),
first != 1 ~ " ") ) %>%
select(-first, -total, -test)
# ta da! ####
confidenceintervalsexp2 <- lrmetable
confidenceintervalsexp2 <- confidenceintervalsexp2 %>%
mutate('OR 95\\% CI' = paste(round(estimate,3)," (",round(conf.low,3),", ",round(conf.high,3),")",sep=""),
'P value' = MoffittFunctions::pretty_pvalues(p.value, background = "yellow")) %>%
mutate(numpvalue = as.numeric(gsub("<","",`Overall P`)),
testOverallP = ifelse(numpvalue < 0.0500001 & is.na(numpvalue) == FALSE, paste("\\cellcolor{yellow}{",`Overall P`,"}",sep=""), `Overall P`),
`Overall P` = testOverallP) %>%
select(Variable, Level, 'OR 95\\% CI','P value','Overall P') %>%
mutate(Variable = gsub("_"," ", Variable))
return(confidenceintervalsexp2)
} #end o'function
globalVariables(c("Overall P","numpvalue","total","testOverallP" ))
#' Fancy Table Output of Mixed effects COX PH model
#'
#' pretty_mixed_effects_coxme() Cox mixed effects model using the coxme() function. Hazard ratios are produced with confidence intervals. P values are also produced. For categorical variables with 3+ levels overall Type 3 p values are calculated, in addition to p values comparing to the first level (reference).
#'
#' @param fit glmer fit
#' @param df is data.frame or tibble used to create model fits. Used for capturing variable labels, if they exist
#' @return
#'
#' A data.frame with: \code{Variable}, \code{Level}, \code{HR (95\% CI)}, \code{P Value} (for categorical variables comparing to reference), \code{Overall P Value} (for categorical variables with 3+ levels).
#'
#' @importFrom lme4 fixef
#' @examples
#' # Mixed effect cox ph model example - random intercepts
#' library(magrittr)
#' library(ggeffects)
#' library(sjmisc)
#' library(lme4)
#' library(splines)
#' library(coxme)
#' library(car)
#' library(survival)
#' library(broom.mixed)
#' library(rlang)
#' library(kableExtra)
#' data(PDACdata)
#' set.seed(542542522)
#' #USE THIS ONE for 3 months to 2 years (24 months)
#'surv_months_obj <- survival::Surv(time = PDACdata$dx_lastcontact_death_months,
#' event = PDACdata$puf_vital_status == 0)
#' cactfit <- coxme(surv_months_obj ~
#' caci3 + surgery_volume_recode + (1|puf_facility_id),
#' data = PDACdata,
#' control=coxme.control(eps = 1e-04, toler.chol = .Machine$double.eps^0.75,
#' iter.max = 200, inner.iter = Quote(max(4, fit0$iter+1)),
#' sparse.calc = NULL,
#' optpar = list(method = "BFGS", control=list(reltol = 1e-3)),
#' refine.df=4, refine.detail=FALSE, refine.method="control",
#' sparse=c(50, .02),
#' varinit=c(.02, .1, .4, .8)^2, corinit = c(0, .3)) )
#' cactfitcmetable <- pretty_mixed_effects_coxme(cactfit, PDACdata)
#' # for PDF output
#' # kable(cactfitcmetable , 'latex', escape = FALSE, longtable = T, booktabs = TRUE, linesep = '',
#' # caption = 'Full Mulitvariable Mixed effects Cox Proportional-Hazards Regression Model with
#' # interaction for Overall Survival2 to 10 years')
#'
#' @export
pretty_mixed_effects_coxme <- function(fit, df){# fit is glmer object, and data frame is the data used.
#code cme table ####
mod<-fit;
extract_coxme_table <- function (mod){
beta <- fixef(mod)
nvar <- length(beta)
nfrail <- nrow(mod$var) - nvar;
# print("the class of mod$var is: ")
# print(class(mod$var))
# print(diag(mod$var)[1:3])
# print(as.numeric(diag(mod$var)[1:3]))
# print(class(diag(mod$var)[nfrail + 1:nvar]))
#
# print(slotNames(mod$var))
# print(class(as.matrix(mod$var )))
se <- sqrt(diag(as.matrix(mod$var )))[nfrail + 1:nvar] #bdsmatrix
z<- round(beta/se, 2)
p<- signif(1 - pchisq((beta/se)^2, 1), 2)
tableout=data.frame(cbind(beta,se,z,p))
return(tableout)
}
reppedvec <- function(fitresults, df){
thevars <- attributes(fitresults$terms)$term.labels;
nvars <- length(attributes(fitresults$terms)$term.labels);
thevars <- gsub("`","", thevars)
therepeats <- rep(NA,nvars)
for (i in 1:nvars){#print(i);
therepeats[i] <- ifelse("numeric" %in% class(df[,thevars[i]]), 1, length(unique(df[,thevars[i]]))- 1)
}
return(therepeats)
}
Variable <- rep(attributes(fit$terms)$term.labels, times = reppedvec(fit, df) )
testout <- cbind(round(cbind(exp( fixef(fit)), exp(confint(fit, level = 0.95))), 3), 'P Value'= extract_coxme_table(fit)$p)
coxmeoutput00 <- as.data.frame(testout) %>% mutate('HR 95\\%' = paste(testout[, 1]," (",testout[, 2],", ",testout[, 3],")",sep=""))
coxmeoutput0 <- cbind(Variable, Level = rownames(testout), coxmeoutput00[, c(5,4)] )
anova(fit) -> fitanova
anovastats <- as.data.frame(cbind(Variable = rownames(fitanova)[-1],
'Overall P' = MoffittFunctions::pretty_pvalues(fitanova$"Pr(>|Chi|)"[-1]) ))
#MERGE in overall pvalue from anova function
cmetable <- as.data.frame( dplyr::left_join(coxmeoutput0, anovastats, by = "Variable") %>%
mutate( 'P Value' = MoffittFunctions::pretty_pvalues(`P Value`, background = "yellow")) %>%
group_by(Variable) %>%
mutate(first = 1:dplyr::n(),
total = dplyr::n()) ) %>%
mutate('Overall P' = case_when(first == 1 & (first != total) ~ as.character(`Overall P`) ,
first == total ~ "",
first != total ~ "" ) ) %>%
ungroup() %>%
mutate(Variable = as.character(Variable),
Level = as.character(Level) )
#remove variable names from levels
test <- rep(NA,NROW(cmetable))
for(i in 1:NROW(cmetable)){
test[i] <- ifelse((cmetable$first[i] == 1 & (cmetable$first[i] == cmetable$total[i])) ,
" ", stringr::str_replace(cmetable[i,2], cmetable[i,1], "") ) }
cmetable$test <- test
cmetable <- cmetable %>%
mutate( Level = test) %>%
mutate( Variable = case_when(first == 1 ~ as.character(Variable),
first != 1 ~ " ") ) %>%
mutate(numpvalue = as.numeric(gsub("<","",`Overall P`)),
testOverallP = ifelse(numpvalue < 0.0500001 & is.na(numpvalue) == FALSE,
paste("\\cellcolor{yellow}{",`Overall P`,"}",sep=""), `Overall P`),
`Overall P` = testOverallP) %>%
select( -first, -total, -test, -numpvalue, -testOverallP) %>%
mutate(Variable = gsub("_"," ", Variable))
return(cmetable)
} # End pretty coxme o'function
z2thet/MoffittFunctions documentation built on July 17, 2021, 9:51 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions pretty_mixed_effects_coxme pretty_mixed_effects_glmer
Documented in pretty_mixed_effects_coxme pretty_mixed_effects_glmer
#' Fancy Table Output of Mixed effects model from glmer
#'
#' pretty_mixed_effects_glmer() takes a mixed effects model from glmer() fit object and calculates estimates, odds ratios with confidence intervals. P values are also produced. For categorical variables with 3+ levels overall Type 3 p values are calculated, in addition to p values comparing to the first level (reference).
#'
#' @param fit glmer fit
#' @param df is data.frame or tibble used to create model fits. Used for capturing variable labels, if they exist
#' @param rdintterm random term in the model
#'
#' @return
#'
#' A data.frame with: \code{Variable}, \code{Level}, \code{ OR (95\% CI)}, \code{P Value} (for categorical variables comparing to reference), \code{Overall P Value} (for categorical variables with 3+ levels).
#'
#' @importFrom dplyr %>%
#' @importFrom dplyr case_when mutate group_by select ungroup
#' @importFrom stats anova
#' @importFrom broom.mixed tidy
#'
#' @examples
#'
#'# glmer fit ####
#' library(magrittr)
#' library(ggeffects)
#' library(sjmisc)
#' library(lme4)
#' library(splines)
#' library(coxme)
#' library(car)
#' library(survival)
#' library(broom.mixed)
#' library(rlang)
#' library(kableExtra)
#' data(PDACdata)
#' gmlmerfit <- glmer(I(PDACdata$mortality90day)=="Dead" ~ caci3
#' + surgery_volume_recode
#' + (1|puf_facility_id)
#' ,family = binomial(link = "logit"),
#' data = PDACdata )
#'
#' # CALL function include fit name, data frame name and random intercept term name in quotes ####
#' outputtable <- pretty_mixed_effects_glmer(gmlmerfit, PDACdata, "puf_facility_id")
#'
#' kableExtra::kable(outputtable, 'html', caption = 'My Table') %>%
#' kableExtra::collapse_rows(c(1:2), row_group_label_position = 'stack')
#'
#' # PDF output
#' #kable(outputtable, 'latex', escape = FALSE, longtable = T, booktabs = TRUE,
#' # linesep = '', caption = 'Multivariable Mixed Effects Logistic Regression
#' # Model Results for 90 day mortality with CACI, Volume, and Facility type')%>%
#' # footnote(number = c('OR are odds of death within 90 days
#' # after the most definitive primary site surgery'))
#'
#'
#' @export
pretty_mixed_effects_glmer <- function(fit,df,rdintterm){# fit is glmer object, and df = data frame is the data used.
#rdintterm = is the name of the variable used for random intercept in quotes
df<-df;
fitresults<-fit;
confidenceintervalsexp2<-tidy(fit,conf.int=TRUE, exponentiate=TRUE, effects="fixed");
car::Anova(fit, test = "Chisq") -> LRfitanova;
LRanovastats <- as.data.frame(cbind(Variable = rownames(LRfitanova),
'Overall P' = MoffittFunctions::pretty_pvalues(LRfitanova$'Pr(>Chisq)') ));
reppedvec <- function(fitresults,df){
facid<-match("puf_facility_id",names(attributes(fit)$frame));
thevars <-names(attributes(fit)$frame)[c(-1,-facid)]#attributes(LRmodel_rint0$terms)$term.labels;
thevars <- gsub("`","", thevars)
nvars <- length(thevars);
therepeats <- rep(NA,nvars)
for (i in 1:nvars){ therepeats[i] <- ifelse("numeric" %in% class(df[,thevars[i]]), 1, length(unique(df[,thevars[i]]))- 1) }
return(therepeats)
}
# NEED match names to get rid of facilty id ####
facid<- match(rdintterm, names(attributes(fit)$frame))
Variable <- rep(names(attributes(fit)$frame)[c(-1,-facid)], times = reppedvec(fit,df ) )
confidenceintervalsexp2$Variable <- c("Intercept",Variable);
lrmetable <- as.data.frame( dplyr::left_join(confidenceintervalsexp2, LRanovastats, by = "Variable") %>%
mutate( 'P Value' = MoffittFunctions::pretty_pvalues(p.value)) %>%
group_by(Variable) %>%
mutate(first = 1:dplyr::n(),
total = dplyr::n()) %>%
mutate(
'Overall P' = case_when(first == 1 & (first != total) ~ as.character(`Overall P`) ,
first == total ~ "",
first != total ~ "" ))%>%
ungroup() %>%
mutate(Variable = as.character(Variable),
Level = as.character(term) )
)#end as data .frame
# remove variable names from levels ####
test <- rep(NA,NROW(lrmetable))
for(i in 1:NROW(lrmetable)){
test[i] <- ifelse((lrmetable$first[i] == 1 & (lrmetable$first[i] == lrmetable$total[i])) ,
" ",stringr::str_replace(lrmetable[i,"Level"], lrmetable[i,"Variable"], "") ) }
lrmetable$test <- test
lrmetable <- lrmetable %>%
mutate( Level = test) %>%
mutate( Variable = case_when(first == 1 ~ as.character(Variable),
first != 1 ~ " ") ) %>%
select(-first, -total, -test)
# ta da! ####
confidenceintervalsexp2 <- lrmetable
confidenceintervalsexp2 <- confidenceintervalsexp2 %>%
mutate('OR 95\\% CI' = paste(round(estimate,3)," (",round(conf.low,3),", ",round(conf.high,3),")",sep=""),
'P value' = MoffittFunctions::pretty_pvalues(p.value, background = "yellow")) %>%
mutate(numpvalue = as.numeric(gsub("<","",`Overall P`)),
testOverallP = ifelse(numpvalue < 0.0500001 & is.na(numpvalue) == FALSE, paste("\\cellcolor{yellow}{",`Overall P`,"}",sep=""), `Overall P`),
`Overall P` = testOverallP) %>%
select(Variable, Level, 'OR 95\\% CI','P value','Overall P') %>%
mutate(Variable = gsub("_"," ", Variable))
return(confidenceintervalsexp2)
} #end o'function
globalVariables(c("Overall P","numpvalue","total","testOverallP" ))
#' Fancy Table Output of Mixed effects COX PH model
#'
#' pretty_mixed_effects_coxme() Cox mixed effects model using the coxme() function. Hazard ratios are produced with confidence intervals. P values are also produced. For categorical variables with 3+ levels overall Type 3 p values are calculated, in addition to p values comparing to the first level (reference).
#'
#' @param fit glmer fit
#' @param df is data.frame or tibble used to create model fits. Used for capturing variable labels, if they exist
#' @return
#'
#' A data.frame with: \code{Variable}, \code{Level}, \code{HR (95\% CI)}, \code{P Value} (for categorical variables comparing to reference), \code{Overall P Value} (for categorical variables with 3+ levels).
#'
#' @importFrom lme4 fixef
#' @examples
#' # Mixed effect cox ph model example - random intercepts
#' library(magrittr)
#' library(ggeffects)
#' library(sjmisc)
#' library(lme4)
#' library(splines)
#' library(coxme)
#' library(car)
#' library(survival)
#' library(broom.mixed)
#' library(rlang)
#' library(kableExtra)
#' data(PDACdata)
#' set.seed(542542522)
#' #USE THIS ONE for 3 months to 2 years (24 months)
#'surv_months_obj <- survival::Surv(time = PDACdata$dx_lastcontact_death_months,
#' event = PDACdata$puf_vital_status == 0)
#' cactfit <- coxme(surv_months_obj ~
#' caci3 + surgery_volume_recode + (1|puf_facility_id),
#' data = PDACdata,
#' control=coxme.control(eps = 1e-04, toler.chol = .Machine$double.eps^0.75,
#' iter.max = 200, inner.iter = Quote(max(4, fit0$iter+1)),
#' sparse.calc = NULL,
#' optpar = list(method = "BFGS", control=list(reltol = 1e-3)),
#' refine.df=4, refine.detail=FALSE, refine.method="control",
#' sparse=c(50, .02),
#' varinit=c(.02, .1, .4, .8)^2, corinit = c(0, .3)) )
#' cactfitcmetable <- pretty_mixed_effects_coxme(cactfit, PDACdata)
#' # for PDF output
#' # kable(cactfitcmetable , 'latex', escape = FALSE, longtable = T, booktabs = TRUE, linesep = '',
#' # caption = 'Full Mulitvariable Mixed effects Cox Proportional-Hazards Regression Model with
#' # interaction for Overall Survival2 to 10 years')
#'
#' @export
pretty_mixed_effects_coxme <- function(fit, df){# fit is glmer object, and data frame is the data used.
#code cme table ####
mod<-fit;
extract_coxme_table <- function (mod){
beta <- fixef(mod)
nvar <- length(beta)
nfrail <- nrow(mod$var) - nvar;
# print("the class of mod$var is: ")
# print(class(mod$var))
# print(diag(mod$var)[1:3])
# print(as.numeric(diag(mod$var)[1:3]))
# print(class(diag(mod$var)[nfrail + 1:nvar]))
#
# print(slotNames(mod$var))
# print(class(as.matrix(mod$var )))
se <- sqrt(diag(as.matrix(mod$var )))[nfrail + 1:nvar] #bdsmatrix
z<- round(beta/se, 2)
p<- signif(1 - pchisq((beta/se)^2, 1), 2)
tableout=data.frame(cbind(beta,se,z,p))
return(tableout)
}
reppedvec <- function(fitresults, df){
thevars <- attributes(fitresults$terms)$term.labels;
nvars <- length(attributes(fitresults$terms)$term.labels);
thevars <- gsub("`","", thevars)
therepeats <- rep(NA,nvars)
for (i in 1:nvars){#print(i);
therepeats[i] <- ifelse("numeric" %in% class(df[,thevars[i]]), 1, length(unique(df[,thevars[i]]))- 1)
}
return(therepeats)
}
Variable <- rep(attributes(fit$terms)$term.labels, times = reppedvec(fit, df) )
testout <- cbind(round(cbind(exp( fixef(fit)), exp(confint(fit, level = 0.95))), 3), 'P Value'= extract_coxme_table(fit)$p)
coxmeoutput00 <- as.data.frame(testout) %>% mutate('HR 95\\%' = paste(testout[, 1]," (",testout[, 2],", ",testout[, 3],")",sep=""))
coxmeoutput0 <- cbind(Variable, Level = rownames(testout), coxmeoutput00[, c(5,4)] )
anova(fit) -> fitanova
anovastats <- as.data.frame(cbind(Variable = rownames(fitanova)[-1],
'Overall P' = MoffittFunctions::pretty_pvalues(fitanova$"Pr(>|Chi|)"[-1]) ))
#MERGE in overall pvalue from anova function
cmetable <- as.data.frame( dplyr::left_join(coxmeoutput0, anovastats, by = "Variable") %>%
mutate( 'P Value' = MoffittFunctions::pretty_pvalues(`P Value`, background = "yellow")) %>%
group_by(Variable) %>%
mutate(first = 1:dplyr::n(),
total = dplyr::n()) ) %>%
mutate('Overall P' = case_when(first == 1 & (first != total) ~ as.character(`Overall P`) ,
first == total ~ "",
first != total ~ "" ) ) %>%
ungroup() %>%
mutate(Variable = as.character(Variable),
Level = as.character(Level) )
#remove variable names from levels
test <- rep(NA,NROW(cmetable))
for(i in 1:NROW(cmetable)){
test[i] <- ifelse((cmetable$first[i] == 1 & (cmetable$first[i] == cmetable$total[i])) ,
" ", stringr::str_replace(cmetable[i,2], cmetable[i,1], "") ) }
cmetable$test <- test
cmetable <- cmetable %>%
mutate( Level = test) %>%
mutate( Variable = case_when(first == 1 ~ as.character(Variable),
first != 1 ~ " ") ) %>%
mutate(numpvalue = as.numeric(gsub("<","",`Overall P`)),
testOverallP = ifelse(numpvalue < 0.0500001 & is.na(numpvalue) == FALSE,
paste("\\cellcolor{yellow}{",`Overall P`,"}",sep=""), `Overall P`),
`Overall P` = testOverallP) %>%
select( -first, -total, -test, -numpvalue, -testOverallP) %>%
mutate(Variable = gsub("_"," ", Variable))
return(cmetable)
} # End pretty coxme o'function
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.