R/GF_twomode_function.R
In MatthewSmith430/gouldfernandez: gouldfernandez
Defines functions
GF_twomode
Documented in
GF_twomode
#' @title Gould-Fernandez Two-Mode Brokerage
#'
#' @description This function extracts the two mode brokerage chains from a two-mode network. Only works for interlocking directorates
#' @param g igraph object.
#' @param event_attribute This needs to be a dataframe with the firm name and the sector attribute
#' @param actor_attribute This needs to a dataframe with the person name and the gender attribute
#' @export
#' @return List - 1.)Brokerage path list 2.) Individual role dataframe list
#' 3.) Total two-mode GF results dataframe.
#' @references Jasny, L. and Lubell, M., 2015. Two-mode brokerage in policy networks. Social Networks, 41, pp.36-47.
GF_twomode<-function(g,event_attribute,actor_attribute){
NET<-g
SECTOR_ID<-event_attribute
colnames(SECTOR_ID)<-c("COMPANY ID","SECTOR")
GENDER_ID<-actor_attribute
colnames(GENDER_ID)<-c("name","gender")
sub<-igraph::make_graph( ~ 1-2-3-4-5)
NET<-igraph::simplify(NET)
H<-igraph::subgraph_isomorphisms(sub, NET)
EL<-list()
for (i in 1:length(H)){
R<-H[[i]]
R2<-as.vector(R)
EL[[i]]<-R2
}
EL2<-do.call(rbind.data.frame, EL)
colnames(EL2)<-c("1","2","3","4","5")
ISO<-list()
for (i in 1:length(EL)){
VEC<-EL[[i]]
subCheck<-igraph::induced_subgraph(NET, VEC)
ISO[[i]]<-igraph::ecount(subCheck)
}
DFedges<-plyr::ldply(ISO,data.frame)
colnames(DFedges)<-"Number of Edges"
DATA<-cbind(EL2,DFedges)
dat<-DATA
dat.sort<-t(apply(dat, 1, sort))
DATA2<- dat[!duplicated(dat.sort),]
#DATA2<-DATA2[DATA2$`Number of Edges` == 2,]
igraph::V(NET)$id<-1:length(igraph::V(NET)$name)
Vert<-igraph::get.data.frame(NET, what = "vertices")
DATA2$`1`<-Vert$name[match(DATA2$`1`,Vert$id)]
DATA2$`2`<-Vert$name[match(DATA2$`2`,Vert$id)]
DATA2$`3`<-Vert$name[match(DATA2$`3`,Vert$id)]
DATA2$`4`<-Vert$name[match(DATA2$`4`,Vert$id)]
DATA2$`5`<-Vert$name[match(DATA2$`5`,Vert$id)]
colnames(DATA2)<-c("Firm1","Director1","Firm2BROKER",
"Director2","Firm3","NoEdges")
DATA2<-as.data.frame(DATA2,stringsAsFactors=FALSE)
DATA2$FIRM1.check<-DATA2$Firm1
DATA2$DIR1.check<-DATA2$Director1
DATA2$FIRM2.check<-DATA2$Firm2BROKER
DATA2$DIR2.check<-DATA2$Director2
DATA2$FIRM3.check<-DATA2$Firm3
DATA2$FIRM1.check<-Vert$type[match(DATA2$FIRM1.check,Vert$name)]
DATA2$DIR1.check<-Vert$type[match(DATA2$DIR1.check,Vert$name)]
DATA2$FIRM2.check<-Vert$type[match(DATA2$FIRM2.check,Vert$name)]
DATA2$DIR2.check<-Vert$type[match(DATA2$DIR2.check,Vert$name)]
DATA2$FIRM3.check<-Vert$type[match(DATA2$FIRM3.check,Vert$name)]
NEWDF<-dplyr::filter(DATA2,DATA2$FIRM1.check==FALSE)
NEWDF<-dplyr::filter(NEWDF,NEWDF$FIRM2.check==FALSE)
NEWDF<-dplyr::filter(NEWDF,NEWDF$FIRM3.check==FALSE)
NEWDF2<-NEWDF[,1:6]
UNF<-unique(unique(NEWDF2$Firm1))
WH_LIST<-list()
for (j in 1:length(UNF)){
FIL<-dplyr::filter(NEWDF2,NEWDF2$Firm1==UNF[[j]])
FIL2<-FIL[FIL$Firm3%in%FIL$Firm2BROKER,]
#FIL2<-dplyr::filter(FIL,FIL$Firm2BROKER==NEWDF2$Firm3[[j]])
#FIL3<-dplyr::filter(FIL,NEWDF2$Firm1==NEWDF2$Firm1[[j]])
#NEWDF3<-anti_join(NEWDF2,FIL2,by="NoEdges")
WH<-which(NEWDF2$Firm1 %in% FIL2$Firm1 &
NEWDF2$Firm2BROKER %in% FIL2$Firm2BROKER &
NEWDF2$Firm3 %in% FIL2$Firm3)
WH_LIST[[j]]<-WH
}
WH_B<-unlist(WH_LIST)
WH_B<-unique(WH_B)
NEWDF2<-NEWDF2[-WH_B,]
NEWDF2$SECTOR1<-NEWDF2$Firm1
NEWDF2$SECTOR2<-NEWDF2$Firm2BROKER
NEWDF2$SECTOR3<-NEWDF2$Firm3
NEWDF2$GENDER1<-NEWDF2$Director1
NEWDF2$GENDER2<-NEWDF2$Director2
NEWDF2$SECTOR1<-SECTOR_ID$SECTOR[match(NEWDF2$Firm1,
SECTOR_ID$`COMPANY ID`)]
NEWDF2$GENDER1<-GENDER_ID$gender[match(NEWDF2$Director1,
GENDER_ID$name)]
NEWDF2$SECTOR2<-SECTOR_ID$SECTOR[match(NEWDF2$Firm2,
SECTOR_ID$`COMPANY ID`)]
NEWDF2$GENDER2<-GENDER_ID$gender[match(NEWDF2$Director2,
GENDER_ID$name)]
NEWDF2$SECTOR3<-SECTOR_ID$SECTOR[match(NEWDF2$Firm3,
SECTOR_ID$`COMPANY ID`)]
#Female - 1, Male - 2
COOR_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==
NEWDF2$SECTOR2&
NEWDF2$SECTOR2==NEWDF2$SECTOR3)
if (length(COOR_DF$Firm1)>0){
COOR_DF$GENDER_COMBO<-rep(0,length(COOR_DF$Firm1))
for (i in 1:length(COOR_DF$GENDER_COMBO)){
CH<-COOR_DF[i,]
Cg1<-CH$GENDER1
Cg2<-CH$GENDER2
if (Cg1==Cg2&&Cg1=="female"){
COOR_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Cg1==Cg2&&Cg1=="male"){
COOR_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Cg1!=Cg2&&Cg1=="male"){
COOR_DF$GENDER_COMBO[[i]]<-"MF"
}else {COOR_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
COOR_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
REP_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==NEWDF2$SECTOR2&
NEWDF2$SECTOR2!=NEWDF2$SECTOR3|NEWDF2$SECTOR2==NEWDF2$SECTOR3&
NEWDF2$SECTOR2!=NEWDF2$SECTOR1)
if (length(REP_DF$Firm1)>0){
REP_DF$GENDER_COMBO<-rep(0,length(REP_DF$Firm1))
for (i in 1:length(REP_DF$GENDER_COMBO)){
RH<-REP_DF[i,]
Rg1<-RH$GENDER1
Rg2<-RH$GENDER2
if (Rg1==Rg2&&Rg1=="female"){
REP_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Rg1==Rg2&&Rg1=="male"){
REP_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Rg1!=Rg2&&Rg1=="male"){
REP_DF$GENDER_COMBO[[i]]<-"MF"
}else {REP_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
REP_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
ITIN_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==NEWDF2$SECTOR3&
NEWDF2$SECTOR1!=NEWDF2$SECTOR2)
if (length(ITIN_DF$Firm1)>0){
ITIN_DF$GENDER_COMBO<-rep(0,length(ITIN_DF$Firm1))
for (i in 1:length(ITIN_DF$GENDER_COMBO)){
IH<-ITIN_DF[i,]
Ig1<-IH$GENDER1
Ig2<-IH$GENDER2
if (Ig1==Ig2&&Ig1=="female"){
ITIN_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Ig1==Ig2&&Ig1=="male"){
ITIN_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Ig1!=Ig2&&Ig1=="male"){
ITIN_DF$GENDER_COMBO[[i]]<-"MF"
}else {ITIN_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
ITIN_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
LIA_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1!=NEWDF2$SECTOR2&
NEWDF2$SECTOR2!=NEWDF2$SECTOR3&NEWDF2$SECTOR1!=NEWDF2$SECTOR3)
if (length(LIA_DF$Firm1)>0){
LIA_DF$GENDER_COMBO<-rep(0,length(LIA_DF$Firm1))
for (i in 1:length(LIA_DF$GENDER_COMBO)){
LH<-LIA_DF[i,]
Lg1<-LH$GENDER1
Lg2<-LH$GENDER2
if (Lg1==Lg2&&Lg1=="female"){
LIA_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Lg1==Lg2&&Lg1=="male"){
LIA_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Lg1!=Lg2&&Lg1=="male"){
LIA_DF$GENDER_COMBO[[i]]<-"MF"
}else {LIA_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
LIA_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
BROKERAGE_PATH_LIST<-list(COOR_DF,REP_DF,ITIN_DF,LIA_DF)
GCL<-c("MM","FF","FM","MF")
##############################################################################
##I need to add an if statement to check whether the DF is empty and alter
if (is.na(REP_DF$Firm1[1])){
REP_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="RepGate")
REP_RES3<-data.frame(Broker=NA,
RepGate.MM=NA,RepGate.FF=NA,RepGate.FM=NA,
RepGate.MF=NA)
REP_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="RepGate")
REP_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,RepGate.MM=NA,
RepGate.FF=NA,RepGate.FM=NA,RepGate.MF=NA)
}else {
REP_RES1<-cbind(REP_DF$Firm2BROKER,REP_DF$GENDER_COMBO
)
REP_RES1<-tibble::as_tibble(REP_RES1)
REP_RES_G1<-cbind(REP_DF$Firm2BROKER,
REP_DF$GENDER_COMBO,REP_DF$SECTOR2
)
REP_RES_G1<-tibble::as_tibble(REP_RES_G1)
REP_RES_G2<-stats::ftable(REP_RES_G1)
REP_RES_G2<-tibble::as_tibble(REP_RES_G2)
colnames(REP_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
REP_RES_G2<-dplyr::filter(REP_RES_G2,Freq!=0)
#REP_RES_G2$Role<-rep("RepGate",length(REP_RES_G2$Freq))
REP_RES_G3<-tidyr::spread(REP_RES_G2,GenderCombo,Freq)
REP_RES2<-table(REP_RES1)
REP_RES2<-tibble::as_tibble(REP_RES2)
colnames(REP_RES2)<-c("Broker","GenderCombo","Freq")
REP_RES2<-dplyr::filter(REP_RES2,Freq!=0)
#REP_RES2$Role<-rep("RepGate",length(REP_RES2$Freq))
REP_RES3<-tidyr::spread(REP_RES2,GenderCombo,Freq)#melt(REP_RES2,id.vars = "Broker")
NR<-GCL[!(GCL%in%unique(REP_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
REP_RES3[,y]<-rep(0,length(REP_RES3$Broker))
REP_RES_G3[,y]<-rep(0,length(REP_RES_G3$Broker))
}
}else{
REP_RES3<-REP_RES3
REP_RES_G3<-REP_RES_G3
}
colnames(REP_RES3)[2:5]<-paste0("RepGate.",colnames(REP_RES3)[2:5])
colnames(REP_RES_G3)[3:6]<-paste0("RepGate.",colnames(REP_RES_G3)[3:6])
}
#############################################################################
if (is.na(COOR_DF$Firm1[1])){
COOR_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Coordinator")
COOR_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Coordinator")
COOR_RES3<-data.frame(Broker=NA,
Coordinator.MM=NA,Coordinator.FF=NA,Coordinator.FM=NA,
Coordinator.MF=NA)
COOR_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Coordinator.MM=NA,
Coordinator.FF=NA,Coordinator.FM=NA,Coordinator.MF=NA)
}else{
COOR_RES1<-cbind(COOR_DF$Firm2BROKER,COOR_DF$GENDER_COMBO
)
COOR_RES1<-tibble::as_tibble(COOR_RES1)
COOR_RES_G1<-cbind(COOR_DF$Firm2BROKER,
COOR_DF$GENDER_COMBO,COOR_DF$SECTOR2
)
COOR_RES_G1<-tibble::as_tibble(COOR_RES_G1)
COOR_RES_G2<-stats::ftable(COOR_RES_G1)
COOR_RES_G2<-tibble::as_tibble(COOR_RES_G2)
colnames(COOR_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
COOR_RES_G2<-dplyr::filter(COOR_RES_G2,Freq!=0)
#COOR_RES_G2$Role<-rep("Coordinator",length(COOR_RES_G2$Freq))
COOR_RES2<-table(COOR_RES1)
COOR_RES2<-tibble::as_tibble(COOR_RES2)
colnames(COOR_RES2)<-c("Broker","GenderCombo","Freq")
COOR_RES2<-dplyr::filter(COOR_RES2,Freq!=0)
#COOR_RES2$Role<-rep("Coordinator",length(COOR_RES2$Freq))
COOR_RES3<-tidyr::spread(COOR_RES2,GenderCombo,Freq)
COOR_RES_G3<-tidyr::spread(COOR_RES_G2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(COOR_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
COOR_RES3[,y]<-rep(0,length(COOR_RES3$Broker))
COOR_RES_G3[,y]<-rep(0,length(COOR_RES_G3$Broker))
}
}else{
COOR_RES3<-COOR_RES3
COOR_RES_G3<-COOR_RES_G3
}
colnames(COOR_RES3)[2:5]<-paste0("Coordinator.",colnames(COOR_RES3)[2:5])
colnames(COOR_RES_G3)[3:6]<-paste0("Coordinator.",colnames(COOR_RES_G3)[3:6])
}
#############################################################################
if (is.na(LIA_DF$Firm1[1])){
LIA_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Liaison")
LIA_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Liaison")
LIA_RES3<-data.frame(Broker=NA,
Liaison.MM=NA,Liaison.FF=NA,Liaison.FM=NA,
Liaison.MF=NA)
LIA_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Liaison.MM=NA,
Liaison.FF=NA,Liaison.FM=NA,Liaison.MF=NA)
}else{
LIA_RES1<-cbind(LIA_DF$Firm2BROKER,LIA_DF$GENDER_COMBO
)
LIA_RES1<-tibble::as_tibble(LIA_RES1)
LIA_RES_G1<-cbind(LIA_DF$Firm2BROKER,
LIA_DF$GENDER_COMBO,LIA_DF$SECTOR2
)
LIA_RES_G1<-tibble::as_tibble(LIA_RES_G1)
LIA_RES_G2<-stats::ftable(LIA_RES_G1)
LIA_RES_G2<-tibble::as_tibble(LIA_RES_G2)
colnames(LIA_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
LIA_RES_G2<-dplyr::filter(LIA_RES_G2,Freq!=0)
#LIA_RES_G2$Role<-rep("Liaison",length(LIA_RES_G2$Freq))
LIA_RES2<-table(LIA_RES1)
LIA_RES2<-tibble::as_tibble(LIA_RES2)
colnames(LIA_RES2)<-c("Broker","GenderCombo","Freq")
LIA_RES2<-dplyr::filter(LIA_RES2,Freq!=0)
#LIA_RES2$Role<-rep("Liaison",length(LIA_RES2$Freq))
LIA_RES_G3<-tidyr::spread(LIA_RES_G2,GenderCombo,Freq)
LIA_RES3<-tidyr::spread(LIA_RES2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(LIA_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
LIA_RES3[,y]<-rep(0,length(LIA_RES3$Broker))
LIA_RES_G3[,y]<-rep(0,length(LIA_RES_G3$Broker))
}
}else{
LIA_RES3<-LIA_RES3
LIA_RES_G3<-LIA_RES_G3
}
colnames(LIA_RES3)[2:5]<-paste0("Liaison.",colnames(LIA_RES3)[2:5])
colnames(LIA_RES_G3)[3:6]<-paste0("Liaison.",colnames(LIA_RES_G3)[3:6])
}
############################################################################
if (is.na(ITIN_DF$Firm1[1])){
ITIN_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Itinerant")
ITIN_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Itinerant")
ITIN_RES3<-data.frame(Broker=NA,
Liaison.MM=NA,Liaison.FF=NA,Liaison.FM=NA,
Liaison.MF=NA)
ITIN_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Liaison.MM=NA,
Liaison.FF=NA,Liaison.FM=NA,Liaison.MF=NA)
}else{
ITIN_RES1<-cbind(ITIN_DF$Firm2BROKER,ITIN_DF$GENDER_COMBO
)
ITIN_RES1<-tibble::as_tibble(ITIN_RES1)
ITIN_RES_G1<-cbind(ITIN_DF$Firm2BROKER,
ITIN_DF$GENDER_COMBO,ITIN_DF$SECTOR2
)
ITIN_RES_G1<-tibble::as_tibble(ITIN_RES_G1)
ITIN_RES_G2<-stats::ftable(ITIN_RES_G1)
ITIN_RES_G2<-tibble::as_tibble(ITIN_RES_G2)
colnames(ITIN_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
ITIN_RES_G2<-dplyr::filter(ITIN_RES_G2,Freq!=0)
#ITIN_RES_G2$Role<-rep("Itinerant",length(ITIN_RES_G2$Freq))
ITIN_RES2<-table(ITIN_RES1)
ITIN_RES2<-tibble::as_tibble(ITIN_RES2)
colnames(ITIN_RES2)<-c("Broker","GenderCombo","Freq")
ITIN_RES2<-dplyr::filter(ITIN_RES2,Freq!=0)
#ITIN_RES2$Role<-rep("Itinerant",length(ITIN_RES2$Freq))
ITIN_RES_G3<-tidyr::spread(ITIN_RES_G2,GenderCombo,Freq)
ITIN_RES3<-tidyr::spread(ITIN_RES2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(ITIN_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
ITIN_RES3[,y]<-rep(0,length(ITIN_RES3$Broker))
ITIN_RES_G3[,y]<-rep(0,length(ITIN_RES_G3$Broker))
}
}else{
ITIN_RES3<-ITIN_RES3
ITIN_RES_G3<-ITIN_RES_G3
}
colnames(ITIN_RES3)[2:5]<-paste0("Itinerant.",colnames(ITIN_RES3)[2:5])
colnames(ITIN_RES_G3)[3:6]<-paste0("Itinerant.",colnames(ITIN_RES_G3)[3:6])
}
################################################################
COM_ID<-dplyr::filter(Vert,Vert$type==FALSE)
COM_ID<-dplyr::select(COM_ID,"name")
CRep<-merge(COM_ID,REP_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CLia<-merge(COM_ID,LIA_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CIter<-merge(COM_ID,ITIN_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CCoor<-merge(COM_ID,COOR_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
TOTAL_DF<-merge(CRep,CLia,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF<-merge(TOTAL_DF,CIter,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF<-merge(TOTAL_DF,CCoor,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF[is.na(TOTAL_DF)]<-0
TOTAL_DF$SECTOR<-TOTAL_DF$name
TOTAL_DF$SECTOR<-SECTOR_ID$SECTOR[match(TOTAL_DF$name,SECTOR_ID$`COMPANY ID`)]
RESULTS_LIST<-list(COOR_RES_G3,REP_RES_G3,ITIN_RES_G3,LIA_RES_G3)
GF_LIST<-list(BROKERAGE_PATH_LIST,RESULTS_LIST,TOTAL_DF)
return(GF_LIST)
}
MatthewSmith430/gouldfernandez documentation built on Dec. 31, 2020, 3:15 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions GF_twomode
Documented in GF_twomode
#' @title Gould-Fernandez Two-Mode Brokerage
#'
#' @description This function extracts the two mode brokerage chains from a two-mode network. Only works for interlocking directorates
#' @param g igraph object.
#' @param event_attribute This needs to be a dataframe with the firm name and the sector attribute
#' @param actor_attribute This needs to a dataframe with the person name and the gender attribute
#' @export
#' @return List - 1.)Brokerage path list 2.) Individual role dataframe list
#' 3.) Total two-mode GF results dataframe.
#' @references Jasny, L. and Lubell, M., 2015. Two-mode brokerage in policy networks. Social Networks, 41, pp.36-47.
GF_twomode<-function(g,event_attribute,actor_attribute){
NET<-g
SECTOR_ID<-event_attribute
colnames(SECTOR_ID)<-c("COMPANY ID","SECTOR")
GENDER_ID<-actor_attribute
colnames(GENDER_ID)<-c("name","gender")
sub<-igraph::make_graph( ~ 1-2-3-4-5)
NET<-igraph::simplify(NET)
H<-igraph::subgraph_isomorphisms(sub, NET)
EL<-list()
for (i in 1:length(H)){
R<-H[[i]]
R2<-as.vector(R)
EL[[i]]<-R2
}
EL2<-do.call(rbind.data.frame, EL)
colnames(EL2)<-c("1","2","3","4","5")
ISO<-list()
for (i in 1:length(EL)){
VEC<-EL[[i]]
subCheck<-igraph::induced_subgraph(NET, VEC)
ISO[[i]]<-igraph::ecount(subCheck)
}
DFedges<-plyr::ldply(ISO,data.frame)
colnames(DFedges)<-"Number of Edges"
DATA<-cbind(EL2,DFedges)
dat<-DATA
dat.sort<-t(apply(dat, 1, sort))
DATA2<- dat[!duplicated(dat.sort),]
#DATA2<-DATA2[DATA2$`Number of Edges` == 2,]
igraph::V(NET)$id<-1:length(igraph::V(NET)$name)
Vert<-igraph::get.data.frame(NET, what = "vertices")
DATA2$`1`<-Vert$name[match(DATA2$`1`,Vert$id)]
DATA2$`2`<-Vert$name[match(DATA2$`2`,Vert$id)]
DATA2$`3`<-Vert$name[match(DATA2$`3`,Vert$id)]
DATA2$`4`<-Vert$name[match(DATA2$`4`,Vert$id)]
DATA2$`5`<-Vert$name[match(DATA2$`5`,Vert$id)]
colnames(DATA2)<-c("Firm1","Director1","Firm2BROKER",
"Director2","Firm3","NoEdges")
DATA2<-as.data.frame(DATA2,stringsAsFactors=FALSE)
DATA2$FIRM1.check<-DATA2$Firm1
DATA2$DIR1.check<-DATA2$Director1
DATA2$FIRM2.check<-DATA2$Firm2BROKER
DATA2$DIR2.check<-DATA2$Director2
DATA2$FIRM3.check<-DATA2$Firm3
DATA2$FIRM1.check<-Vert$type[match(DATA2$FIRM1.check,Vert$name)]
DATA2$DIR1.check<-Vert$type[match(DATA2$DIR1.check,Vert$name)]
DATA2$FIRM2.check<-Vert$type[match(DATA2$FIRM2.check,Vert$name)]
DATA2$DIR2.check<-Vert$type[match(DATA2$DIR2.check,Vert$name)]
DATA2$FIRM3.check<-Vert$type[match(DATA2$FIRM3.check,Vert$name)]
NEWDF<-dplyr::filter(DATA2,DATA2$FIRM1.check==FALSE)
NEWDF<-dplyr::filter(NEWDF,NEWDF$FIRM2.check==FALSE)
NEWDF<-dplyr::filter(NEWDF,NEWDF$FIRM3.check==FALSE)
NEWDF2<-NEWDF[,1:6]
UNF<-unique(unique(NEWDF2$Firm1))
WH_LIST<-list()
for (j in 1:length(UNF)){
FIL<-dplyr::filter(NEWDF2,NEWDF2$Firm1==UNF[[j]])
FIL2<-FIL[FIL$Firm3%in%FIL$Firm2BROKER,]
#FIL2<-dplyr::filter(FIL,FIL$Firm2BROKER==NEWDF2$Firm3[[j]])
#FIL3<-dplyr::filter(FIL,NEWDF2$Firm1==NEWDF2$Firm1[[j]])
#NEWDF3<-anti_join(NEWDF2,FIL2,by="NoEdges")
WH<-which(NEWDF2$Firm1 %in% FIL2$Firm1 &
NEWDF2$Firm2BROKER %in% FIL2$Firm2BROKER &
NEWDF2$Firm3 %in% FIL2$Firm3)
WH_LIST[[j]]<-WH
}
WH_B<-unlist(WH_LIST)
WH_B<-unique(WH_B)
NEWDF2<-NEWDF2[-WH_B,]
NEWDF2$SECTOR1<-NEWDF2$Firm1
NEWDF2$SECTOR2<-NEWDF2$Firm2BROKER
NEWDF2$SECTOR3<-NEWDF2$Firm3
NEWDF2$GENDER1<-NEWDF2$Director1
NEWDF2$GENDER2<-NEWDF2$Director2
NEWDF2$SECTOR1<-SECTOR_ID$SECTOR[match(NEWDF2$Firm1,
SECTOR_ID$`COMPANY ID`)]
NEWDF2$GENDER1<-GENDER_ID$gender[match(NEWDF2$Director1,
GENDER_ID$name)]
NEWDF2$SECTOR2<-SECTOR_ID$SECTOR[match(NEWDF2$Firm2,
SECTOR_ID$`COMPANY ID`)]
NEWDF2$GENDER2<-GENDER_ID$gender[match(NEWDF2$Director2,
GENDER_ID$name)]
NEWDF2$SECTOR3<-SECTOR_ID$SECTOR[match(NEWDF2$Firm3,
SECTOR_ID$`COMPANY ID`)]
#Female - 1, Male - 2
COOR_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==
NEWDF2$SECTOR2&
NEWDF2$SECTOR2==NEWDF2$SECTOR3)
if (length(COOR_DF$Firm1)>0){
COOR_DF$GENDER_COMBO<-rep(0,length(COOR_DF$Firm1))
for (i in 1:length(COOR_DF$GENDER_COMBO)){
CH<-COOR_DF[i,]
Cg1<-CH$GENDER1
Cg2<-CH$GENDER2
if (Cg1==Cg2&&Cg1=="female"){
COOR_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Cg1==Cg2&&Cg1=="male"){
COOR_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Cg1!=Cg2&&Cg1=="male"){
COOR_DF$GENDER_COMBO[[i]]<-"MF"
}else {COOR_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
COOR_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
REP_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==NEWDF2$SECTOR2&
NEWDF2$SECTOR2!=NEWDF2$SECTOR3|NEWDF2$SECTOR2==NEWDF2$SECTOR3&
NEWDF2$SECTOR2!=NEWDF2$SECTOR1)
if (length(REP_DF$Firm1)>0){
REP_DF$GENDER_COMBO<-rep(0,length(REP_DF$Firm1))
for (i in 1:length(REP_DF$GENDER_COMBO)){
RH<-REP_DF[i,]
Rg1<-RH$GENDER1
Rg2<-RH$GENDER2
if (Rg1==Rg2&&Rg1=="female"){
REP_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Rg1==Rg2&&Rg1=="male"){
REP_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Rg1!=Rg2&&Rg1=="male"){
REP_DF$GENDER_COMBO[[i]]<-"MF"
}else {REP_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
REP_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
ITIN_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1==NEWDF2$SECTOR3&
NEWDF2$SECTOR1!=NEWDF2$SECTOR2)
if (length(ITIN_DF$Firm1)>0){
ITIN_DF$GENDER_COMBO<-rep(0,length(ITIN_DF$Firm1))
for (i in 1:length(ITIN_DF$GENDER_COMBO)){
IH<-ITIN_DF[i,]
Ig1<-IH$GENDER1
Ig2<-IH$GENDER2
if (Ig1==Ig2&&Ig1=="female"){
ITIN_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Ig1==Ig2&&Ig1=="male"){
ITIN_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Ig1!=Ig2&&Ig1=="male"){
ITIN_DF$GENDER_COMBO[[i]]<-"MF"
}else {ITIN_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
ITIN_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
LIA_DF<-dplyr::filter(NEWDF2,NEWDF2$SECTOR1!=NEWDF2$SECTOR2&
NEWDF2$SECTOR2!=NEWDF2$SECTOR3&NEWDF2$SECTOR1!=NEWDF2$SECTOR3)
if (length(LIA_DF$Firm1)>0){
LIA_DF$GENDER_COMBO<-rep(0,length(LIA_DF$Firm1))
for (i in 1:length(LIA_DF$GENDER_COMBO)){
LH<-LIA_DF[i,]
Lg1<-LH$GENDER1
Lg2<-LH$GENDER2
if (Lg1==Lg2&&Lg1=="female"){
LIA_DF$GENDER_COMBO[[i]]<-"FF"
}else if(Lg1==Lg2&&Lg1=="male"){
LIA_DF$GENDER_COMBO[[i]]<-"MM"
} else if (Lg1!=Lg2&&Lg1=="male"){
LIA_DF$GENDER_COMBO[[i]]<-"MF"
}else {LIA_DF$GENDER_COMBO[[i]]<-"FM"}
}
}else{
LIA_DF<-data.frame(Firm1=NA,
Director1=NA,
Firm2BROKER=NA,
Firm3=NA,NoEdges=NA,
SECTOR1=NA,SECTOR2=NA,SECTOR3=NA,
GENDER1=NA,GENDER2=NA,
GENDER_COMBO=NA)
}
BROKERAGE_PATH_LIST<-list(COOR_DF,REP_DF,ITIN_DF,LIA_DF)
GCL<-c("MM","FF","FM","MF")
##############################################################################
##I need to add an if statement to check whether the DF is empty and alter
if (is.na(REP_DF$Firm1[1])){
REP_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="RepGate")
REP_RES3<-data.frame(Broker=NA,
RepGate.MM=NA,RepGate.FF=NA,RepGate.FM=NA,
RepGate.MF=NA)
REP_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="RepGate")
REP_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,RepGate.MM=NA,
RepGate.FF=NA,RepGate.FM=NA,RepGate.MF=NA)
}else {
REP_RES1<-cbind(REP_DF$Firm2BROKER,REP_DF$GENDER_COMBO
)
REP_RES1<-tibble::as_tibble(REP_RES1)
REP_RES_G1<-cbind(REP_DF$Firm2BROKER,
REP_DF$GENDER_COMBO,REP_DF$SECTOR2
)
REP_RES_G1<-tibble::as_tibble(REP_RES_G1)
REP_RES_G2<-stats::ftable(REP_RES_G1)
REP_RES_G2<-tibble::as_tibble(REP_RES_G2)
colnames(REP_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
REP_RES_G2<-dplyr::filter(REP_RES_G2,Freq!=0)
#REP_RES_G2$Role<-rep("RepGate",length(REP_RES_G2$Freq))
REP_RES_G3<-tidyr::spread(REP_RES_G2,GenderCombo,Freq)
REP_RES2<-table(REP_RES1)
REP_RES2<-tibble::as_tibble(REP_RES2)
colnames(REP_RES2)<-c("Broker","GenderCombo","Freq")
REP_RES2<-dplyr::filter(REP_RES2,Freq!=0)
#REP_RES2$Role<-rep("RepGate",length(REP_RES2$Freq))
REP_RES3<-tidyr::spread(REP_RES2,GenderCombo,Freq)#melt(REP_RES2,id.vars = "Broker")
NR<-GCL[!(GCL%in%unique(REP_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
REP_RES3[,y]<-rep(0,length(REP_RES3$Broker))
REP_RES_G3[,y]<-rep(0,length(REP_RES_G3$Broker))
}
}else{
REP_RES3<-REP_RES3
REP_RES_G3<-REP_RES_G3
}
colnames(REP_RES3)[2:5]<-paste0("RepGate.",colnames(REP_RES3)[2:5])
colnames(REP_RES_G3)[3:6]<-paste0("RepGate.",colnames(REP_RES_G3)[3:6])
}
#############################################################################
if (is.na(COOR_DF$Firm1[1])){
COOR_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Coordinator")
COOR_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Coordinator")
COOR_RES3<-data.frame(Broker=NA,
Coordinator.MM=NA,Coordinator.FF=NA,Coordinator.FM=NA,
Coordinator.MF=NA)
COOR_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Coordinator.MM=NA,
Coordinator.FF=NA,Coordinator.FM=NA,Coordinator.MF=NA)
}else{
COOR_RES1<-cbind(COOR_DF$Firm2BROKER,COOR_DF$GENDER_COMBO
)
COOR_RES1<-tibble::as_tibble(COOR_RES1)
COOR_RES_G1<-cbind(COOR_DF$Firm2BROKER,
COOR_DF$GENDER_COMBO,COOR_DF$SECTOR2
)
COOR_RES_G1<-tibble::as_tibble(COOR_RES_G1)
COOR_RES_G2<-stats::ftable(COOR_RES_G1)
COOR_RES_G2<-tibble::as_tibble(COOR_RES_G2)
colnames(COOR_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
COOR_RES_G2<-dplyr::filter(COOR_RES_G2,Freq!=0)
#COOR_RES_G2$Role<-rep("Coordinator",length(COOR_RES_G2$Freq))
COOR_RES2<-table(COOR_RES1)
COOR_RES2<-tibble::as_tibble(COOR_RES2)
colnames(COOR_RES2)<-c("Broker","GenderCombo","Freq")
COOR_RES2<-dplyr::filter(COOR_RES2,Freq!=0)
#COOR_RES2$Role<-rep("Coordinator",length(COOR_RES2$Freq))
COOR_RES3<-tidyr::spread(COOR_RES2,GenderCombo,Freq)
COOR_RES_G3<-tidyr::spread(COOR_RES_G2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(COOR_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
COOR_RES3[,y]<-rep(0,length(COOR_RES3$Broker))
COOR_RES_G3[,y]<-rep(0,length(COOR_RES_G3$Broker))
}
}else{
COOR_RES3<-COOR_RES3
COOR_RES_G3<-COOR_RES_G3
}
colnames(COOR_RES3)[2:5]<-paste0("Coordinator.",colnames(COOR_RES3)[2:5])
colnames(COOR_RES_G3)[3:6]<-paste0("Coordinator.",colnames(COOR_RES_G3)[3:6])
}
#############################################################################
if (is.na(LIA_DF$Firm1[1])){
LIA_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Liaison")
LIA_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Liaison")
LIA_RES3<-data.frame(Broker=NA,
Liaison.MM=NA,Liaison.FF=NA,Liaison.FM=NA,
Liaison.MF=NA)
LIA_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Liaison.MM=NA,
Liaison.FF=NA,Liaison.FM=NA,Liaison.MF=NA)
}else{
LIA_RES1<-cbind(LIA_DF$Firm2BROKER,LIA_DF$GENDER_COMBO
)
LIA_RES1<-tibble::as_tibble(LIA_RES1)
LIA_RES_G1<-cbind(LIA_DF$Firm2BROKER,
LIA_DF$GENDER_COMBO,LIA_DF$SECTOR2
)
LIA_RES_G1<-tibble::as_tibble(LIA_RES_G1)
LIA_RES_G2<-stats::ftable(LIA_RES_G1)
LIA_RES_G2<-tibble::as_tibble(LIA_RES_G2)
colnames(LIA_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
LIA_RES_G2<-dplyr::filter(LIA_RES_G2,Freq!=0)
#LIA_RES_G2$Role<-rep("Liaison",length(LIA_RES_G2$Freq))
LIA_RES2<-table(LIA_RES1)
LIA_RES2<-tibble::as_tibble(LIA_RES2)
colnames(LIA_RES2)<-c("Broker","GenderCombo","Freq")
LIA_RES2<-dplyr::filter(LIA_RES2,Freq!=0)
#LIA_RES2$Role<-rep("Liaison",length(LIA_RES2$Freq))
LIA_RES_G3<-tidyr::spread(LIA_RES_G2,GenderCombo,Freq)
LIA_RES3<-tidyr::spread(LIA_RES2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(LIA_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
LIA_RES3[,y]<-rep(0,length(LIA_RES3$Broker))
LIA_RES_G3[,y]<-rep(0,length(LIA_RES_G3$Broker))
}
}else{
LIA_RES3<-LIA_RES3
LIA_RES_G3<-LIA_RES_G3
}
colnames(LIA_RES3)[2:5]<-paste0("Liaison.",colnames(LIA_RES3)[2:5])
colnames(LIA_RES_G3)[3:6]<-paste0("Liaison.",colnames(LIA_RES_G3)[3:6])
}
############################################################################
if (is.na(ITIN_DF$Firm1[1])){
ITIN_RES2<-data.frame(Broker=NA,
GenderCombo=NA,
Freq=NA,
Role="Itinerant")
ITIN_RES_G2<-data.frame(Broker=NA,
GenderCombo=NA,
BrokerSector=NA,
Freq=NA,
Role="Itinerant")
ITIN_RES3<-data.frame(Broker=NA,
Liaison.MM=NA,Liaison.FF=NA,Liaison.FM=NA,
Liaison.MF=NA)
ITIN_RES_G3<-data.frame(Broker=NA,BrokerSector=NA,Liaison.MM=NA,
Liaison.FF=NA,Liaison.FM=NA,Liaison.MF=NA)
}else{
ITIN_RES1<-cbind(ITIN_DF$Firm2BROKER,ITIN_DF$GENDER_COMBO
)
ITIN_RES1<-tibble::as_tibble(ITIN_RES1)
ITIN_RES_G1<-cbind(ITIN_DF$Firm2BROKER,
ITIN_DF$GENDER_COMBO,ITIN_DF$SECTOR2
)
ITIN_RES_G1<-tibble::as_tibble(ITIN_RES_G1)
ITIN_RES_G2<-stats::ftable(ITIN_RES_G1)
ITIN_RES_G2<-tibble::as_tibble(ITIN_RES_G2)
colnames(ITIN_RES_G2)<-c("Broker","GenderCombo","BrokerSector","Freq")
ITIN_RES_G2<-dplyr::filter(ITIN_RES_G2,Freq!=0)
#ITIN_RES_G2$Role<-rep("Itinerant",length(ITIN_RES_G2$Freq))
ITIN_RES2<-table(ITIN_RES1)
ITIN_RES2<-tibble::as_tibble(ITIN_RES2)
colnames(ITIN_RES2)<-c("Broker","GenderCombo","Freq")
ITIN_RES2<-dplyr::filter(ITIN_RES2,Freq!=0)
#ITIN_RES2$Role<-rep("Itinerant",length(ITIN_RES2$Freq))
ITIN_RES_G3<-tidyr::spread(ITIN_RES_G2,GenderCombo,Freq)
ITIN_RES3<-tidyr::spread(ITIN_RES2,GenderCombo,Freq)
NR<-GCL[!(GCL%in%unique(ITIN_RES2$GenderCombo))]
if (length(NR)>0){
for (i in 1:length(NR)){
y<-NR[i]
ITIN_RES3[,y]<-rep(0,length(ITIN_RES3$Broker))
ITIN_RES_G3[,y]<-rep(0,length(ITIN_RES_G3$Broker))
}
}else{
ITIN_RES3<-ITIN_RES3
ITIN_RES_G3<-ITIN_RES_G3
}
colnames(ITIN_RES3)[2:5]<-paste0("Itinerant.",colnames(ITIN_RES3)[2:5])
colnames(ITIN_RES_G3)[3:6]<-paste0("Itinerant.",colnames(ITIN_RES_G3)[3:6])
}
################################################################
COM_ID<-dplyr::filter(Vert,Vert$type==FALSE)
COM_ID<-dplyr::select(COM_ID,"name")
CRep<-merge(COM_ID,REP_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CLia<-merge(COM_ID,LIA_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CIter<-merge(COM_ID,ITIN_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
CCoor<-merge(COM_ID,COOR_RES3,by.x="name",
by.y="Broker",all.x=TRUE)
TOTAL_DF<-merge(CRep,CLia,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF<-merge(TOTAL_DF,CIter,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF<-merge(TOTAL_DF,CCoor,
by.x = "name",by.y="name",all.x=TRUE)
TOTAL_DF[is.na(TOTAL_DF)]<-0
TOTAL_DF$SECTOR<-TOTAL_DF$name
TOTAL_DF$SECTOR<-SECTOR_ID$SECTOR[match(TOTAL_DF$name,SECTOR_ID$`COMPANY ID`)]
RESULTS_LIST<-list(COOR_RES_G3,REP_RES_G3,ITIN_RES_G3,LIA_RES_G3)
GF_LIST<-list(BROKERAGE_PATH_LIST,RESULTS_LIST,TOTAL_DF)
return(GF_LIST)
}
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.