OB2_exploratory_Rscript.Rmd

R script for:
Transgenerational effects of maternal obesogenic diet on appetite: a meta-analysis  
========================================================

Step 2 - Exploratory analyses  
--------------------

```{r start, echo=FALSE, message=FALSE, eval=TRUE, include=FALSE}
options(scipen=100)

#getRversion() #3.1.1

install.packages(c("ProjectTemplate", "knitr", "MCMCglmm", "metafor", "rmeta", "meta", "xtable",  "arm", "ade4", "psych","orthopolynom"))

library('ProjectTemplate')

#create.project('R_OB_analyses')#use once to initiate

load.project()

#project.info$data

library(xtable)
library(meta)
library(psych)
library(lattice)

#library(metafor)
#library(MCMCglmm)

load("data_NC.RData") 
load("data.RData") 
```

Preprocessed data based on he raw data: raw_main_data_20140815_OB.csv and raw_diet_data_20140815_OB.csv

Preprocessing done so far:  
* Removed records where dams were on both nutrient manipulation and caloric restriction.  
* Removed records for offspring on self-selected diets (altered food preferences could bias total intake).  
* Aggregated time series-type data by ES_ID.  
* Merged with diet information dataframe, for dam and offspring diets.  
* Calculated effect sizes for 4 outcomes: BW, AI, RI, RI2.  


```{r data summary, echo=FALSE, eval=TRUE, include=FALSE}
length(unique(data$Study_ID)) # 53 studies (was 27 before update)
length(unique(data$Exp_ID)) # 61 experments
table(data$Offspr_sex) # F 35, FM 12, M 69
table(data$animal) #mouse 27, rat 89
table(data$Strain) 
# C57BL/6       C57BL/6J            FVB            ICR     Long-Evans           NMRI Sprague-Dawley         Wistar 
#     5             12              2              6              6              2             46             37 
table(data$Dam_choice_diet) # 69 - 0, 47 - 1
table(data$Dam_diet_lactation_incl) # 23 gestation only, 93 gestation+lactation
table(data$Dam_choice_diet,data$Dam_diet_lactation_incl) # 10, 59, 13, 34 - ok
table(data$Litter_size_equalized) #  7 NA,  12 no, 97 yes 
```

Studies included in the final obesogenic dataset: total `r nrow(data)` ES (was 57 ES before update) from `r length(unique(data$Study_ID))` studies (was 27), representing `r length(unique(data$Exp_ID))`separate experiments.  

* By sex: `r table(data$Offspr_sex)`  
* By species: `r table(data$animal)`  
* By strain: `r table(data$Strain)`  
* By dam choice diet: `r table(data$Dam_choice_diet)`  
* By lactation included in treatment: `r table(data$Dam_diet_lactation_incl)`  

Table of ES by species/strain  
-------------------------------

```{r strains table, results='asis', echo=FALSE}

### STRAINS TABLE - collapse dataset to the strain level
#table(data$Study_ID, data$Strain)
countStrains <- sapply(tapply(data$ES_ID, data$Strain, unique), length) 
#countS #4 strains: C57BL/6J - 2, Long-Evans - 4, Sprague-Dawley - 26, Wistar - 25
tabStrains <- cbind(data[match(names(countStrains),data$Strain),"animal"],countStrains)
print(xtable(tabStrains), type='html') #display in markdown
#kable(xtable(tabStrains), format = "markdown") #test generic function in kntr1.5
write.csv(tabStrains,"OB_strains_table.csv") #save as .csv
#write.table(print.xtable(xtable(tabStrains),type="html",caption.placement="top"),file="OB_strains_table.html") #save as .html
```

Table of included studies  
----------------------------

```{r studies table, results='asis', echo=FALSE, eval=TRUE, include=FALSE}
### PAPERS TABLE - collapse dataset the to paper level
countES <- sapply(tapply(data$ES_ID, data$Study_ID, unique), length) 
countES #1-4 ES per study, usually 1 or 2
tab <- cbind(data[match(names(countES),data$Study_ID),c("Study_ID","First_author","Year","animal","Strain")],countES) 
tab
#table(data$Study_ID,data$Offspr_sex)  #check if diff sexes from the same study reported - sometimes yes
#sapply(tapply(data$Exp_ID, data$Study_ID, unique), length) #1-4 experiments per study, but usually 1
#sapply(tapply(data$ES_ID, data$Exp_ID, unique), length) #1-4 ES per experiment, but usually 1

#aggregate by Study ID concatenating values for "Original_measur_units","Dam_choice_diet","Offspr_choice_diet","Offspr_sex":
tab2 <- cbind(tab, aggregate(Original_measur_units ~ Study_ID, data = data, FUN=function(x) paste(unique(x), collapse = ","))[,2], aggregate(Dam_choice_diet ~ Study_ID, data = data, FUN=function(x) paste(unique(x), collapse = ","))[,2], aggregate(Offspr_sex ~ Study_ID, data = data, FUN=function(x) paste(unique(x), collapse = ","))[,2])
names(tab2) <- c(names(tab2)[1:3],"Species","Strain","N of ES","Original_measur_units","Dam_choice_diet","Offspr_sex") #rename columns
tab2 <- tab2[order(tab2$Species,tab2$Strain), ] #order by species and strain
rownames(tab2) <- NULL
print(xtable(tab2), type='html') #display in markdown
#kable(xtable(tab2), format = "markdown") #test generic function in kntr1.5
write.csv(tab2,"OB_papers_table.csv") #save as .csv
#write.table(print.xtable(xtable(tab2),type="html",caption.placement="top"),file="OB_papers_table.html") #save as .html

#ES with missing data on dam diet composition
dim(data[which(is.na(data$Dam_diet_exp_E)),]) #24 effect sizes lack experimental diet composition data (these are all choice diets) out of 46 ES with choice diets (average composition extracted)
#Note:  should not use diet composition from choice diets as covariate in models, as many values missing or very crude!
```

Papers table - second version
Include: Species,  Strain,	Reference,	Source of data for effect size, diet type, N E.S.
```{r studies table2, results='asis', echo=FALSE, eval=TRUE, include=FALSE}

tab3 <- cbind(data[match(names(countES),data$Study_ID),c("First_author","Year","Study_ID","animal","Strain","Dam_choice_diet","Offspr_data_source")],countES) 
tab3 <- tab3[!is.na(tab3$Year),] #remove empty rows
tab3$Reference <- paste(tab3$First_author,tab3$Year) #paste togather First author and year
tab3 <- tab3[,-(1:3)] #remove first 3 columns
tab3$Dam_choice_diet <- ifelse(tab3$Dam_choice_diet==1,"choice","chow")
names(tab3) <- c("Species","Strain","Dam diet type","Data source","N of ES","Reference") #rename columns
tab3 <- tab3[order(tab3$Species,tab3$Strain), ] #order by species and strain
rownames(tab3) <- NULL

#print(xtable(tab3), type='html') #display in markdown
#kable(xtable(tab3), format = "markdown") #test generic function in kntr1.5
write.csv(tab3,"OB_papers_table2.csv") #save as .csv
#write.table(print.xtable(xtable(tab3),type="html",caption.placement="top"),file="OB_papers_table2.html") #save as .html

```
*******

Explore diet properties in the chow (no-choice) data subset  
----------------------------------------------------------
Summary of the chow diet subset.  
```{r chow diets subset, echo=FALSE, eval=TRUE, include=FALSE}
#create chow diet (no-choice) data subset
data_chow <- data[data$Dam_choice_diet == "0",] #69 ES
#data_chow <- subset(data_chow, Dam_choice_diet=="0")

table(data_chow$Offspr_sex) # F 19, FM 1, M 47 - exclude FM as only 1 ES left!
length(unique(data_chow$Study_ID)) # 31 studies
length(unique(data_chow$Exp_ID)) # 38 experments
table(data_chow$Offspr_sex) # F 19, FM 1, M 49
table(data_chow$animal) #mouse 20, rat 49
table(data_chow$Strain) #8 strains
table(data_chow$Dam_diet_lactation_incl) # 10 gestation only, 59 gestation+lactation
#data_chow68 <- subset(data_chow, Offspr_sex!="FM") #68 ES - for analyses with offspring sex moderator!
#write.csv(data_chow,"NM_overfeeding_NCsubset.csv")
#data_chow <- read.csv("NM_overfeeding_NCsubset.csv")
```

Studies included in the final obesogenic chow diet data subset: total `r nrow(data_chow)` ES from `r length(unique(data_chow$Study_ID))` studies, representing `r length(unique(data_chow$Exp_ID))`separate experiments.  

* By sex: `r table(data_chow$Offspr_sex)`  
* By species: `r table(data_chow$animal)`  
* By strain: `r table(data_chow$Strain)`  
* By lactation included in treatment: `r table(data_chow$Dam_diet_lactation_incl)`  


Explore diets in chow diet data subset  
```{r chow diets summary, echo=FALSE, eval=TRUE, include=FALSE}
#fat content in experimental diet
dim(data_chow[which(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE > 10),]) #65 out of 69 ES use increased fat content (10% more energy in treatment group than in control group)
dim(data_chow[which(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE < -10),]) #2 out of 69 ES use decreased fat content (10% more energy in treatment group than in control group)

dim(data_chow[which(data_chow$Dam_diet_exp_carbE - data_chow$Dam_diet_contr_carbE > 10),]) #2 out of 69 ES use increased carbs content (10% more energy in treatment group than in control group)
dim(data_chow[which(data_chow$Dam_diet_exp_carbE - data_chow$Dam_diet_contr_carbE < -10),]) #65 out of 69 ES use decreased carbs content (10% more energy in treatment group than in control group)

dim(data_chow[which(data_chow$Dam_diet_exp_protE - data_chow$Dam_diet_contr_protE > 10),]) #0 out of 69 ES use increased prots content (10% more energy in treatment group than in control group)
dim(data_chow[which(data_chow$Dam_diet_exp_protE - data_chow$Dam_diet_contr_protE < -10),]) #11 out of 69 ES use decreased prots content (10% more energy in treatment group than in control group)


hist(data_chow$Dam_diet_contr_fatE) #fat usually 10-15%E
hist(data_chow$Dam_diet_contr_carbE) #carbs usually 50-75%E
hist(data_chow$Dam_diet_contr_protE) #prots usually 15-30%E
hist(data_chow$Dam_diet_exp_fatE) #fat usually 40-70%E
hist(data_chow$Dam_diet_exp_carbE) #carbs usually 10-40%E
hist(data_chow$Dam_diet_exp_protE) #prots usually 15-22%E
hist(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE) #usually fat increased by approx. 30%E
hist(data_chow$Dam_diet_exp_carbE - data_chow$Dam_diet_contr_carbE) #usually carbs reduced by approx. 30%E
hist(data_chow$Dam_diet_exp_protE - data_chow$Dam_diet_contr_protE) #protein content same or reduced by approx 10%E 


#Kozak 2000, 2005, Couvreur 2011 - high carb diets, lower fats in exp. dam groups - remove for sensitivity analyses
dim(data_chow[which(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE <= 10),]) #4 out of 69 ES do not use increased fat content 
data_chow[which(data_chow$Dam_diet_exp_E - data_chow$Dam_diet_contr_E < 0),] #2 also have E lower in exp. group than in contr. dams diets
data_chow[which(data_chow$Dam_diet_exp_fatE > 70),] #2 ES (Kozak 2000, Kozak 2005) have more than 70% of energy from fat 

#check dam-offspring diet mismatch (included only offspring on show diets!)
dim(data_chow[which(data_chow$Offspr_diet_E - data_chow$Dam_diet_contr_E == 0),]) #47 ES offspring on same E diet as control dams
dim(data_chow[which(data_chow$Offspr_diet_E - data_chow$Dam_diet_contr_E < 0),]) #8 ES offspring on lower E diet than control dams
dim(data_chow[which(data_chow$Offspr_diet_E - data_chow$Dam_diet_contr_E <= 0),]) #55 ES offspring on same or lower E diet than control dams (47+8)

dim(data_chow[which(data_chow$Offspr_diet_E - data_chow$Dam_diet_contr_E > 0),]) #14 ES offspring on higher E diet than control dams
dim(data_chow[which(data_chow$Offspr_diet_E - data_chow$Dam_diet_exp_E == 0),]) #11 ES offspring on same E diet as exp dams (from 14 above)

#make new column - whether offspring was on obesogenic diet (higher E than control dams)
#data_chow$Offspr_diet_OB <- (data_chow$Offspr_diet_E - data_chow$Dam_diet_contr_E > 0) #19 TRUE #line moved to preprocessing script!
```

Out of `r nrow(data_chow)` ES, `r dim(data_chow[which(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE > 10),])[1]` used high-fat diets, `r dim(data_chow[which(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE <= 10),])[1]` high-carb diets. No missing data.  

# Plot dam and offspring diets characteristics in the chow (no-choice) data set.
```{r chow diets subset hist, fig.width=8, fig.height=11, echo=FALSE, eval=TRUE, include=FALSE}
#pdf(file="graphs/Rplot_chow_diets_hist.pdf",width=9,height=11,family="sans",pointsize=12)
par(mfcol=c(5,4),mar=c(5,4,1,1.75),oma=c(1,1,0,0))
#control dam diets
hist(data_chow$Dam_diet_contr_E, xlim=c(2,8), ylim=c(0,60), breaks=seq(2,8,0.5), main="Control dam diets", xlab="total energy [kcal/g]") #E usually 3-4 kcal/g
hist(data_chow$Dam_diet_contr_fatE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="fats [%E]") #fat usually 10-15%E
hist(data_chow$Dam_diet_contr_carbE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="carbs [%E]") #carbs usually 50-75%E
hist(data_chow$Dam_diet_contr_protE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="proteins [%E]") #prots usually 15-30%E
hist(data_chow$Dam_diet_contr_PNP_kcal, xlim=c(0,1), ylim=c(0,60), breaks=seq(0,1,0.05), main="", xlab="P:NP ratio") #PNP usually 
#experimental dam diets
hist(data_chow$Dam_diet_exp_E, xlim=c(2,8), ylim=c(0,60), breaks=seq(2,8,0.5), main="Experimental dam diets", xlab="total energy [kcal/g]", ylab="") #E usually 4-6 kcal/g
hist(data_chow$Dam_diet_exp_fatE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="fats [%E]", ylab="") #fat usually 40-70%E
hist(data_chow$Dam_diet_exp_carbE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="carbs [%E]", ylab="") #carbs usually 10-40%E
hist(data_chow$Dam_diet_exp_protE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="proteins [%E]", ylab="") #prots usually 15-22%E
hist(data_chow$Dam_diet_exp_PNP_kcal, xlim=c(0,1), ylim=c(0,60), breaks=seq(0,1,0.05), main="", xlab="P:NP ratio", ylab="") #PNP usually 
#difference exp-control dam diets
hist(data_chow$Dam_diet_exp_E - data_chow$Dam_diet_contr_E, xlim=c(-4,4), ylim=c(0,60), breaks=seq(-4,4,0.5), main="Control-Experimental", xlab="total energy [kcal/g]", ylab="") #E usually increased by 1-2 kcl/g
hist(data_chow$Dam_diet_exp_fatE - data_chow$Dam_diet_contr_fatE, xlim=c(-60,60), ylim=c(0,60), breaks=seq(-60,60,10), main="", xlab="fats [%E]", ylab="") #usually fat increased by approx. 30%E
hist(data_chow$Dam_diet_exp_carbE - data_chow$Dam_diet_contr_carbE, xlim=c(-60,60), ylim=c(0,60), breaks=seq(-60,60,10), main="", xlab="carbs [%E]", ylab="") #usually carbs reduced by approx. 30%E
hist(data_chow$Dam_diet_exp_protE - data_chow$Dam_diet_contr_protE, xlim=c(-60,60), ylim=c(0,60), breaks=seq(-60,60,10), main="", xlab="proteins [%E]", ylab="") #protein content same or reduced by approx 10%E 
hist(data_chow$Dam_diet_exp_PNP_kcal - data_chow$Dam_diet_contr_PNP_kcal, xlim=c(-1,1), ylim=c(0,60), breaks=seq(-1,1,0.1), main="", xlab="P:NP ratio", ylab="") #PNP usually 
#offspring diets
hist(data_chow$Offspr_diet_E, xlim=c(2,8), ylim=c(0,60), breaks=seq(2,8,0.5), main="Offspring diets", xlab="total energy [kcal/g]", ylab="") #E usually 3-4 kcal/g
hist(data_chow$Offspr_diet_fatE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="fats [%E]", ylab="") #fat usually 10-15%E
hist(data_chow$Offspr_diet_carbE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="carbs [%E]", ylab="") #carbs usually 50-75%E
hist(data_chow$Offspr_diet_protE, xlim=c(0,100), ylim=c(0,60), breaks=seq(0,100,10), main="", xlab="proteins [%E]", ylab="") #prots usually 15-30%E
hist(data_chow$Offspr_diet_PNP_kcal, xlim=c(0,1), ylim=c(0,60), breaks=seq(0,1,0.05), main="", xlab="P:NP ratio", ylab="") #PNP usually 
#dev.off()
```

# Plot offspring diets characteristics in the full data set.
```{r offspr chow diets all data , fig.width=8, fig.height=6, echo=FALSE}
#vectors of unique control,experimental and offspring diet codes
#c.diets <- as.vector(unique(data$Dam_diet_code_contr)) # 28
#e.diets <- as.vector(unique(data$Dam_diet_code_exp)) # 33
#o.diets <- as.vector(unique(data$Offspr_diet_code)) # 35

#pdf(file="graphs/Rplot_dam_offspr_diets_hist.pdf",width=8,height=6,family="sans",pointsize=10)
par(mfrow=c(3,2),mar=c(5,4,1,1.75),oma=c(1,1,0,0))
hist(data$Offspr_diet_E, xlim=c(2,8), ylim=c(0,80), breaks=seq(2,8,0.5), main="Offspring diet - full data set", xlab="total energy [kcal/g]")
hist(data$Offspr_diet_fatE, xlim=c(0,100), ylim=c(0,80), breaks=seq(0,100,10), main="", xlab="fats [%E]") 
hist(data$Offspr_diet_carbE, xlim=c(0,100), ylim=c(0,80), breaks=seq(0,100,10), main="", xlab="carbs [%E]") 
hist(data$Offspr_diet_protE, xlim=c(0,100), ylim=c(0,80), breaks=seq(0,100,10), main="", xlab="proteins [%E]") 
hist(data$Offspr_diet_PNP_kcal, xlim=c(0,1), ylim=c(0,80), breaks=seq(0,1,0.05), main="", xlab="P:NP ratio")
#dev.off()
```
Mostly same as control diets, but few on obesogenic diets (consider excluding for senesitivity analysis).


Dam diet properties arrow plots:  
--------------------------
# 2 panel plot for dams only (use in the main text) for the chow (no-choice) subset.

```{r diets arrow 2 plots, fig.width=8, fig.height=9, echo=FALSE, eval=TRUE, include=TRUE}

#pdf(file="Rplot_dam_diets_arrows_main.pdf",width=7,height=8,family="sans", pointsize=10)
#pdf(file="Rplot_dam_diets_arrows_main.pdf",paper="a4",family="sans",pointsize=8)

par(mfrow=c(2,1),mar=c(4,4,1,1.75),oma=c(1,1,0,0))

### PLOT A
#plot dam diets as total energy vs. PNP_ratio, with arrows linking control and experimental dams
plot(data_chow$Dam_diet_contr_PNP_kcal,data_chow$Dam_diet_contr_E,xlim=c(0,1),ylim=c(2,7),xlab="diet P:NP [kcal/kcal]",ylab="diet energy [kcal/g]") #control diets
points(data_chow$Dam_diet_exp_PNP_kcal,data_chow$Dam_diet_exp_E,pch=2 ) # experimental diets 
arrows(data_chow$Dam_diet_contr_PNP_kcal,data_chow$Dam_diet_contr_E,data_chow$Dam_diet_exp_PNP_kcal,data_chow$Dam_diet_exp_E,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=0.45, bty="n")
text(0, 7, "A", cex=1.5)
# INTERPRETATION: horizontal arrows - isocaloric protein manipulation diets, other arrows - both calories and protein change (mostly obesogenic diets).

### PLOT B
#plot dam diets as percent energy from fat vs. percent energy from carbs, with arrows linking control and experimental dams 
plot(data_chow$Dam_diet_contr_fatE,data_chow$Dam_diet_contr_carbE,xlim=c(0,100),ylim=c(0,100),xlab="energy from fat [%]",ylab="energy from carbs [%]") #control diets
points(data_chow$Dam_diet_exp_fatE,data_chow$Dam_diet_exp_carbE,pch=2) # experimental diets
arrows(data_chow$Dam_diet_contr_fatE,data_chow$Dam_diet_contr_carbE,data_chow$Dam_diet_exp_fatE,data_chow$Dam_diet_exp_carbE,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=35, bty="n")
text(0, 100, "B", cex=1.5)
#INTERPRETATION: vertical arrows - energy from fat stays the same, but carbs content changes (isocaloric protein manipulation diets), diagonal arrows - obesogenic diets: fat contribution increases and carbs decreases.

#dev.off()
```
A. Almost all experiemntal diets have higher energy content than control dam diets. usually PNP ratio reeduced as result. Variation in energy and PNP within control diets.
B. Experimental diets have more energy from fat than control dam diets. This usually is accompanied by reduction of carbohydrates in the diet.

# 4 panel plot for dams only (dont use)
```{r diets arrow 4 plots, fig.width=8, fig.height=9, echo=FALSE, eval=TRUE, include=FALSE}

#pdf(file="Rplot_dam_diets_arrows.pdf",width=7,height=8,family="sans", pointsize=10)
#pdf(file="Rplot_dam_diets_arrows.pdf",paper="a4",family="sans",pointsize=8)

par(mfrow=c(2,2),mar=c(4,4,1,1.75),oma=c(1,1,0,0))

### PLOT 1
#plot dam diets as total energy vs. PNP_ratio, with arrows linking control and experimental dams
plot(data$Dam_diet_contr_PNP_kcal,data$Dam_diet_contr_E,xlim=c(0,1),ylim=c(2,7),xlab="diet P:NP [kcal/kcal]",ylab="diet energy [kcal/g]") #control diets
points(data$Dam_diet_exp_PNP_kcal,data$Dam_diet_exp_E,pch=2, col=data$Dam_choice_diet+1) # experimental diets (0+1=black=no choice, 1+1=red=choice)
arrows(data$Dam_diet_contr_PNP_kcal,data$Dam_diet_contr_E,data$Dam_diet_exp_PNP_kcal,data$Dam_diet_exp_E,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=0.45, bty="n")
# INTERPRETATION: horizontal arrows - isocaloric protein manipulation diets, other arrows - both calories and protein change (mostly obesogenic diets). Red = choice diets.

### PLOT 2
#plot dam diets as percent energy from protein vs. percent energy from fat, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_protE,data$Dam_diet_contr_fatE,xlim=c(0,80),ylim=c(0,100),xlab="energy from protein [%]",ylab="energy from fat [%]") #control diets
points(data$Dam_diet_exp_protE,data$Dam_diet_exp_fatE,pch=2, col=data$Dam_choice_diet+1) # experimental diets (0+1=black=no choice, 1+1=red=choice)
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_fatE,data$Dam_diet_exp_protE,data$Dam_diet_exp_fatE,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=35, bty="n")
#INTERPRETATION: horizontal arrows - isocaloric protein manipulation diets (fats kept constant), vertical arrows - fat content manipulation (obesogenic diets) without changing protein content in the diet. Red = choice diets.

### PLOT 3
#plot dam diets as percent energy from protein vs. percent energy from carbs, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_protE,data$Dam_diet_contr_carbE,xlim=c(0,100),ylim=c(0,100),xlab="energy from protein [%]",ylab="energy from carbs [%]") #control diets
points(data$Dam_diet_exp_protE,data$Dam_diet_exp_carbE,pch=2, col=data$Dam_choice_diet+1) # experimental diets (0+1=black=no choice, 1+1=red=choice)
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_carbE,data$Dam_diet_exp_protE,data$Dam_diet_exp_carbE,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=35, bty="n")
#INTERPRETATION: diagonal arrows - isocaloric protein manipulation diets done by changing carbs in the diet, vertical arrows - fat content manipulation (obesogenic diets) alsop by changing carbs in the diet, but not protein. Red = choice diets.

### PLOT 4
#plot dam diets as percent energy from fat vs. percent energy from carbs, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_fatE,data$Dam_diet_contr_carbE,xlim=c(0,100),ylim=c(0,100),xlab="energy from fat [%]",ylab="energy from carbs [%]") #control diets
points(data$Dam_diet_exp_fatE,data$Dam_diet_exp_carbE,pch=2, col=data$Dam_choice_diet+1) # experimental diets (0+1=black=no choice, 1+1=red=choice)
arrows(data$Dam_diet_contr_fatE,data$Dam_diet_contr_carbE,data$Dam_diet_exp_fatE,data$Dam_diet_exp_carbE,length=0.1,lwd=0.2) #change in energy and PNP
legend("top",legend=c("control dams","experimental dams"), pch=c(1,2),text.width=35, bty="n")
#INTERPRETATION: vertical arrows - energy from fat stays the same, but carbs content changes (isocaloric protein manipulation diets), diagonal arrows - obesogenic diets: fat contribution increases and carbs decreases. Red = choice diets.

#dev.off()
```

# 4 panel plot for dams and offspring (dont use)
```{r diets plot dam offspr, fig.width=8, fig.height=9, echo=FALSE, warning=FALSE, eval=TRUE, include=FALSE}
#pdf(file="graphs/Rplot_dam_offspr_diets_arrows.pdf",width=7,height=8,family="sans",pointsize=10)
par(mfrow=c(2,2),mar=c(4,4,1,1.75),oma=c(1,1,0,0))

### PLOT 1
#plot dam diets as total energy vs. PNP_ratio, with arrows linking control and experimental dams
plot(data$Dam_diet_contr_PNP_kcal,data$Dam_diet_contr_E,xlim=c(0,1),ylim=c(2,7),xlab="diet P:NP [kcal/kcal]",ylab="diet energy [kcal/g]") #control diets
points(data$Dam_diet_exp_PNP_kcal,data$Dam_diet_exp_E,pch=2) # experimental diets
arrows(data$Dam_diet_contr_PNP_kcal,data$Dam_diet_contr_E,data$Dam_diet_exp_PNP_kcal,data$Dam_diet_exp_E,length=0.1) #change in energy and PNP contr.- exp dams
#legend("topright",legend=c("control dams","experimental dams"), pch=c(1,2), text.width=0.45, bty="n", cex=1)
# INTERPRETATION: ~vertical arrows - both calories and protein change (mostly obesogenic diets).
# CHECK: in few diets total E decreases?
points(data$Offspr_diet_PNP_kcal,data$Offspr_diet_E,pch=3) #energy and PNP in offspring diets
arrows(data$Dam_diet_contr_PNP_kcal,data$Dam_diet_contr_E,data$Offspr_diet_PNP_kcal,data$Offspr_diet_E,length=0.1,lty=3) #change between contr. dams and offspring
arrows(data$Dam_diet_exp_PNP_kcal,data$Dam_diet_exp_E,data$Offspr_diet_PNP_kcal,data$Offspr_diet_E,length=0.1,lty=3) #change between exp. dams and offspring
#NOTE: some offspring diets unquantified (self-selection), others vary widely in composition - use as moderators?
legend("topright",legend=c("control dams","experimental dams","offspring"), pch=c(1,2,3),text.width=0.45, bty="n", cex=1)

### PLOT 2
#plot dam diets as percent energy from protein vs. percent energy from fat, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_protE,data$Dam_diet_contr_fatE,xlim=c(0,80),ylim=c(0,100),xlab="energy from protein [%]",ylab="energy from datat [%]") #control diets
points(data$Dam_diet_exp_protE,data$Dam_diet_exp_fatE,pch=2) # experimental diets
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_fatE,data$Dam_diet_exp_protE,data$Dam_diet_exp_fatE,length=0.1) #change in energy and PNP
#legend("topright",legend=c("control dams","experimental dams"), pch=c(1,2), text.width=35, bty="n", cex=1)
#INTERPRETATION: vertical arrows - fat content manipulation (obesogenic diets) without changing protein content in the diet
points(data$Offspr_diet_protE,data$Offspr_diet_fatE,pch=3) #energy and PNP in offspring diets
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_fatE,data$Offspr_diet_protE,data$Offspr_diet_fatE,length=0.1,lty=3) #change between contr. dams and offspring
arrows(data$Dam_diet_exp_protE,data$Dam_diet_exp_fatE,data$Offspr_diet_protE,data$Offspr_diet_fatE,length=0.1,lty=3) #change between exp. dams and offspring
#NOTE: some offspring diets unquantified (self-selection), others vary widely in composition - use as moderators?
legend("topright",legend=c("control dams","experimental dams","offspring"), pch=c(1,2,3),text.width=35, bty="n", cex=1)

### PLOT 3
#plot dam diets as percent energy from protein vs. percent energy from carbs, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_protE,data$Dam_diet_contr_carbE,xlim=c(0,80),ylim=c(0,100),xlab="energy from protein [%]",ylab="energy from carbs [%]") #control diets
points(data$Dam_diet_exp_protE,data$Dam_diet_exp_carbE,pch=2) # experimental diets
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_carbE,data$Dam_diet_exp_protE,data$Dam_diet_exp_carbE,length=0.1) #change in energy and PNP
#legend("topright",legend=c("control dams","experimental dams"), pch=c(1,2), text.width=35, bty="n", cex=1)
#INTERPRETATION: vertical arrows - fat content manipulation (obesogenic diets) alsop by changing carbs in the diet, but not protein.
points(data$Offspr_diet_protE,data$Offspr_diet_carbE,pch=3) #energy and PNP in offspring diets
arrows(data$Dam_diet_contr_protE,data$Dam_diet_contr_carbE,data$Offspr_diet_protE,data$Offspr_diet_carbE,length=0.1,lty=3) #change between contr. dams and offspring
arrows(data$Dam_diet_exp_protE,data$Dam_diet_exp_carbE,data$Offspr_diet_protE,data$Offspr_diet_carbE,length=0.1,lty=3) #change between exp. dams and offspring
#NOTE: some offspring diets unquantified (self-selection), others vary widely in composition - use as moderators?
legend("topright",legend=c("control dams","experimental dams","offspring"), pch=c(1,2,3),text.width=35, bty="n", cex=1)

### PLOT 4
#plot dam diets as percent energy from fat vs. percent energy from carbs, with arrows linking control and experimental dams 
plot(data$Dam_diet_contr_fatE,data$Dam_diet_contr_carbE,xlim=c(0,80),ylim=c(0,100),xlab="energy from fat [%]",ylab="energy from carbs [%]") #control diets
points(data$Dam_diet_exp_fatE,data$Dam_diet_exp_carbE,pch=2) # experimental diets
arrows(data$Dam_diet_contr_fatE,data$Dam_diet_contr_carbE,data$Dam_diet_exp_fatE,data$Dam_diet_exp_carbE,length=0.1) #change in energy and PNP
#legend("topright",legend=c("control dams","experimental dams"), pch=c(1,2), text.width=35, bty="n", cex=1)
#INTERPRETATION: diagonal arrows - obesogenic diets: fat contribution increases and carbs decreases.
points(data$Offspr_diet_fatE,data$Offspr_diet_carbE,pch=3) #energy and PNP in offspring diets
arrows(data$Dam_diet_contr_fatE,data$Dam_diet_contr_carbE,data$Offspr_diet_fatE,data$Offspr_diet_carbE,length=0.1,lty=3) #change between contr. dams and offspring
arrows(data$Dam_diet_exp_fatE,data$Dam_diet_exp_carbE,data$Offspr_diet_fatE,data$Offspr_diet_carbE,length=0.1,lty=3) #change between exp. dams and offspring
#NOTE: some offspring diets unquantified (self-selection), others vary widely in composition - use as moderators?
legend("topright",legend=c("control dams","experimental dams","offspring"), pch=c(1,2,3),text.width=35, bty="n", cex=1)

#dev.off()
```
  
Set transparent colors for plotting:
```{r transparent colors function, echo=FALSE, eval=TRUE, include=FALSE}
## Add an alpha value to a colour - a function from: http://lamages.blogspot.co.nz/2013/04/how-to-change-alpha-value-of-colours-in.html
# my version is stored in library folder and loaded automatically as a helper function!

#set colors for offspring sex categories: F, FM, M - currently in "helpers"" script
#SexColours <- c("red", "steelblue", "grey")
#SexColoursAlpha <- add.alpha(SexColours, alpha=0.4) #use in plotting functions as: col=SexColoursAlpha["colorname"]
```

FUNNEL PLOTS - full data set
----------------------------

```{r raw funnel plots, fig.width=8, fig.height=9, echo=FALSE, eval=TRUE, include=FALSE}

#pdf(file="Rplot_raw_funnels.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(2,2),mar=c(4,4,1,1.75),oma=c(1,1,0,0))

### PLOT Bodyweight
metaBW_fa <- metagen(data$Bodyweight_Hd, sqrt(data$Bodyweight_Hd_Var))
funnel.meta(metaBW_fa, yaxis = "invse", xlab="Hedge's d", ylab="Precision (1/SE)", xlim=c(-10,10), ylim=c(0,6), pch=ifelse(data$Offspr_sex=="F",16, ifelse(data$Offspr_sex=="M",17,15)), level=NULL, comb.fixed=FALSE, comb.random=FALSE, cex=1.2, col = ifelse(data$Offspr_sex=="F",SexColoursAlpha["red"], ifelse(data$Offspr_sex=="M",SexColoursAlpha["steelblue"],SexColoursAlpha["grey"])))
abline(v=0, lwd=1, lty=2)
title(main = "(a) Body mass [g]", adj=0)
legend("topright",legend=c("females","males","females+males"), pch=c(16,17,15), col=c(SexColoursAlpha["red"],SexColoursAlpha["steelblue"],SexColoursAlpha["grey"]), text.width=7, bty="n", cex=1)

### PLOT AbsIntake
metaAI_fa <- metagen(data$AbsIntake_Hd, sqrt(data$AbsIntake_Hd_Var))
funnel.meta(metaAI_fa, yaxis = "invse", xlab="Hedge's d", ylab="Precision (1/SE)", xlim=c(-10,10), ylim=c(0,6), pch=ifelse(data$Offspr_sex=="F",16, ifelse(data$Offspr_sex=="M",17,15)), level=NULL, comb.fixed=FALSE, comb.random=FALSE, cex=1.2, col = ifelse(data$Offspr_sex=="F",SexColoursAlpha["red"], ifelse(data$Offspr_sex=="M",SexColoursAlpha["steelblue"],SexColoursAlpha["grey"])))
abline(v=0, lwd=1, lty=2)
title(main = "(b) Absolute intake [kcal/day]", adj=0)
legend("topright",legend=c("females","males","females+males"), pch=c(16,17,15), col=c(SexColoursAlpha["red"],SexColoursAlpha["steelblue"],SexColoursAlpha["grey"]), text.width=7, bty="n", cex=1)

### PLOT RelIntake
metaRI_fa <- metagen(data$RelIntake_Hd, sqrt(data$RelIntake_Hd_Var))
funnel.meta(metaRI_fa, yaxis = "invse", xlab="Hedge's d", ylab="Precision (1/SE)", xlim=c(-10,10), ylim=c(0,6), pch=ifelse(data$Offspr_sex=="F",16, ifelse(data$Offspr_sex=="M",17,15)), level=NULL, comb.fixed=FALSE, comb.random=FALSE, cex=1.2, col = ifelse(data$Offspr_sex=="F",SexColoursAlpha["red"], ifelse(data$Offspr_sex=="M",SexColoursAlpha["steelblue"],SexColoursAlpha["grey"])))
abline(v=0, lwd=1, lty=2)
title(main = "(c) Linear intake [kcal/g/day]", adj=0)
legend("topright",legend=c("females","males","females+males"), pch=c(16,17,15), col=c(SexColoursAlpha["red"],SexColoursAlpha["steelblue"],SexColoursAlpha["grey"]), text.width=7, bty="n", cex=1)

### PLOT RelIntake2
metaRI2_fa <- metagen(data$RelIntake2_Hd, sqrt(data$RelIntake2_Hd_Var))
funnel.meta(metaRI2_fa, yaxis = "invse", xlab="Hedge's d", ylab="Precision (1/SE)", xlim=c(-10,10), ylim=c(0,6), pch=ifelse(data$Offspr_sex=="F",16, ifelse(data$Offspr_sex=="M",17,15)), level=NULL, comb.fixed=FALSE, comb.random=FALSE, cex=1.2, col = ifelse(data$Offspr_sex=="F",SexColoursAlpha["red"], ifelse(data$Offspr_sex=="M",SexColoursAlpha["steelblue"],SexColoursAlpha["grey"])))
abline(v=0, lwd=1, lty=2)
title(main = "(d) Allometric intake [kcal/g^0.75/day]", adj=0)
legend("topright",legend=c("females","males","females+males"), pch=c(16,17,15), col=c(SexColoursAlpha["red"],SexColoursAlpha["steelblue"],SexColoursAlpha["grey"]), text.width=7, bty="n", cex=1)

#dev.off()
```

Correlations among outcomes  
---------------------------
Checking the relationship between the 4 outcomes: AI, RI, RI2 and BW  
```{r corr outcomes plot, fig.width=8, fig.height=6, echo=FALSE, eval=TRUE, include=FALSE}
#pdf(file="Rplot_corr_4outcomes.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(1,3),bty="l",lwd=1)

cor.test(data$Bodyweight_Hd,data$AbsIntake_Hd,na.action=exclude) # sign. positive corr.
cor.test(rank(data$Bodyweight_Hd),rank(data$AbsIntake_Hd),na.action=exclude) # sign. positive corr.
plot(data$Bodyweight_Hd,data$AbsIntake_Hd) #bigger diff in BW - bigger diff. in AI (if exp. bigger than cont. they also eat more in total)
abline(lm(data$Bodyweight_Hd~data$AbsIntake_Hd))

cor.test(data$Bodyweight_Hd,data$RelIntake_Hd) # sign. negative corr.
cor.test(rank(data$Bodyweight_Hd),rank(data$RelIntake_Hd)) # sign. negative corr.
plot(data$Bodyweight_Hd,data$RelIntake_Hd) #bigger diff in BW - less diff. in RI (if exp. bigger than cont. they eat less per g bw)
abline(lm(data$Bodyweight_Hd~data$RelIntake_Hd))

cor.test(data$Bodyweight_Hd,data$RelIntake2_Hd) # non-sign. negative corr.
cor.test(rank(data$Bodyweight_Hd),rank(data$RelIntake2_Hd)) # non-sign. negative corr.
plot(data$Bodyweight_Hd,data$RelIntake2_Hd) #no relationship between diff in BW and diff. in RI2, which means RI2 is well adjusted for diff. in BW!
abline(lm(data$Bodyweight_Hd~data$RelIntake2_Hd))
#dev.off()
```
When offspring of experimental dams are bigger than offspring of control dams, thay also eat significantly more (absolute intake d - positove correlation with bodyweight d).
Negative relatonship with linear relative intake - when big difference in body size the big experimental animals seem to eat less per unit of body weight than controls, suggesting allometric relationship.
RelIntake2 (allometric) does not significantly correlate with body mass, which suggests it is well adjusted for differences in body mass.  
  
Correlations among predictors  
-----------------------------

```{r cont mod scatterplot, fig.width=8, fig.height=8, echo=FALSE, eval=TRUE, include=FALSE}
# histograms and correlation scatterplot matrix for raw continuous moderators

hist(data$Intake_age_dPC)
summary(data$Intake_age_dPC)

library(psych)

#names(data)
par(mfrow=c(1,1),bty="l",lwd=1)

#all:
pairs.panels(~Dam_diet_start_exp_dPC+Dam_diet_end_exp_dPC+Dam_diet_duration+Intake_age_dPC+Intake_duration+Dam_diet_contr_E+Dam_diet_contr_PC+Dam_diet_contr_PNP_kcal+Dam_diet_exp_E+Dam_diet_exp_PC+Dam_diet_exp_PNP_kcal+Offspr_diet_E+Offspr_diet_PC+Offspr_diet_PNP_kcal+Year, main = "Correlations between selected continuous predictors",lm=TRUE, ellipses=FALSE, jiggle=TRUE,data=data)

data[data$Dam_diet_contr_E > 4,] #11 ES have energy-rich control diets!

#Maternal diet manip. timing:
pairs.panels(~data$Dam_diet_start_exp_dPC + data$Dam_diet_end_exp_dPC + data$Dam_diet_duration) #start and end dates uncorrelated
dim(data[data$Dam_diet_end_exp_dPC > 22,])#82 ES gestation+lactation
nrow(data[data$Dam_diet_end_exp_dPC <= 22,])#34 ES where gestation-only manipulation
dim(data[data$Dam_diet_start_exp_dPC < 0,])#81 ES with pre-conception start
dim(data[data$Dam_diet_start_exp_dPC >= 0,])#35 ES with pre-conception start

#Maternal diet properties:
Dam_diet_diff_E <- data$Dam_diet_exp_E - data$Dam_diet_contr_E
pairs.panels(~data$Dam_diet_contr_E + data$Dam_diet_exp_E + Dam_diet_diff_E) #contr. corr with exp., difference correlated with exp. but not contr.

Dam_diet_diff_PC < -data$Dam_diet_exp_PC - data$Dam_diet_contr_PC
pairs.panels(~data$Dam_diet_contr_PC + data$Dam_diet_exp_PC + Dam_diet_diff_PC) #contr. NOT corr with exp., difference correlated with exp. but not contr.

Dam_diet_diff_PNP_kcal <- data$Dam_diet_exp_PNP_kcal - data$Dam_diet_contr_PNP_kcal
pairs.panels(~data$Dam_diet_contr_PNP_kcal + data$Dam_diet_exp_PNP_kcal + Dam_diet_diff_PNP_kcal) #contr NOT corr with exp., difference negativ. correlated with contr. (if control had high PNP, it was more likely to decrease in experimental diet!)

#Offspring diet properties:
Dam_diet_diff_E <- data$Dam_diet_exp_E - data$Dam_diet_contr_E
pairs.panels(~data$Offspr_diet_E + data$Offspr_diet_PC + data$Offspr_diet_PNP_kcal) #E some pos. corr. with PC ratio?
```


Exploring categorical moderators - plots
----------------------------------------

```{r categ mod plots, echo=FALSE, eval=TRUE, include=FALSE}

#pdf(file="Rplot_boxplot_Hd_strain_sex.pdf",width=7,height=8,family="sans",pointsize=10)
par(mfcol=c(4,2))

#split by strain
plot(Bodyweight_Hd~Strain,data=data)
plot(AbsIntake_Hd~Strain,data=data)
plot(RelIntake_Hd~Strain,data=data)
plot(RelIntake2_Hd~Strain,data=data)

#split by sex
plot(Bodyweight_Hd~Offspr_sex,data=data)
plot(AbsIntake_Hd~Offspr_sex,data=data)
plot(RelIntake_Hd~Offspr_sex,data=data)
plot(RelIntake2_Hd~Offspr_sex,data=data)

#dev.off()


#pdf(file="Rplot_boxplots_Hd_choice_lactation_mismatch.pdf",width=7,height=8,family="sans",pointsize=10)
par(mfcol=c(4,3))

#split by dam choice
boxplot(Bodyweight_Hd~Dam_choice_diet,data=data,main=("Bodyweight_Hd~Dam_choice_diet"))
boxplot(AbsIntake_Hd~Dam_choice_diet,data=data,main=("AbsIntake_Hd~Dam_choice_diet"))
boxplot(RelIntake_Hd~Dam_choice_diet,data=data,main=("RelIntake_Hd~Dam_choice_diet"))
boxplot(RelIntake2_Hd~Dam_choice_diet,data=data,main=("RelIntake2_Hd~Dam_choice_diet"))

#split by Dam_diet_lactation_incl
boxplot(Bodyweight_Hd~Dam_diet_lactation_incl,data=data,main=("Bodyweight_Hd~Dam_diet_lactation_incl"))
boxplot(AbsIntake_Hd~Dam_diet_lactation_incl,data=data,main=("AbsIntake_Hd~Dam_diet_lactation_incl"))
boxplot(RelIntake_Hd~Dam_diet_lactation_incl,data=data,main=("RelIntake_Hd~Dam_diet_lactation_incl"))
boxplot(RelIntake2_Hd~Dam_diet_lactation_incl,data=data,main=("RelIntake2_Hd~Dam_diet_lactation_incl"))

#split by offspring diet type
boxplot(Bodyweight_Hd~Offspr_diet_OB,data=data,main=("Bodyweight_Hd~Offspr_diet_OB"))
boxplot(AbsIntake_Hd~Offspr_diet_OB,data=data,main=("AbsIntake_Hd~Offspr_diet_OB"))
boxplot(RelIntake_Hd~Offspr_diet_OB,data=data,main=("RelIntake_Hd~Offspr_diet_OB"))
boxplot(RelIntake2_Hd~Offspr_diet_OB,data=data,main=("RelIntake2_Hd~Offspr_diet_OB"))

#dev.off()
```

Bubble plots of dam diet effect by sex
--------------------------------------

```{r bubble plots by sex, echo=FALSE, eval=TRUE, include=FALSE}
###### Bubble plots of effects of selected continuous predictors

#! use data$ for the full data set and fas$ for the non-choice subset!

#######Dam and offspring diet parameters - full data set
### Bodyweight

#pdf(file="Rplot_bubbl_BW.pdf",width=8,height=12,family="sans",pointsize=10)

par(mfrow=c(8,4),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#Bodyweight vs. Dam_diet_exp_E
symbols(data$Dam_diet_exp_E,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",main="Bodyweight - ALL",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="Bodyweight - F",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="Bodyweight - M",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="Bodyweight - FM",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_exp_PNP_kcal
symbols(data$Dam_diet_exp_PNP_kcal,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_contr_E
symbols(data$Dam_diet_contr_E,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_contr_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_contr_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_contr_PNP_kcal
symbols(data$Dam_diet_contr_PNP_kcal,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_contr_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. data$Dam_diet_diff_E
symbols(data$Dam_diet_diff_E,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. data$Dam_diet_diff_PNP
symbols(data$Dam_diet_diff_PNP,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP",ylab="Effect size (Hedge's d)",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_diff_PNP, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Offspr_diet_E
symbols(data$Offspr_diet_E,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Offspr_diet_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Offspr_diet_E, na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Offspr_diet_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Offspr_diet_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Offspr_diet_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Offspr_diet_PNP_kcal
symbols(data$Offspr_diet_PNP_kcal,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Offspr_diet_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)

#dev.off()


### RelIntake2
#pdf(file="Rplot_bubbl_RI2.pdf",width=8,height=12,family="sans",pointsize=10)

par(mfrow=c(8,4),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#RelIntake2 vs. Dam_diet_exp_E
symbols(data$Dam_diet_exp_E,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",main="RelIntake2 - ALL",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="RelIntake2 - F",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="RelIntake2 - M",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_E[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="RelIntake2 - FM",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_exp_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_exp_PNP_kcal
symbols(data$Dam_diet_exp_PNP_kcal,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_exp_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_contr_E
symbols(data$Dam_diet_contr_E,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_contr_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_contr_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_E[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_contr_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_contr_PNP_kcal
symbols(data$Dam_diet_contr_PNP_kcal,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_contr_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_contr_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. data$Dam_diet_diff_E
symbols(data$Dam_diet_diff_E,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. data$Dam_diet_diff_PNP
symbols(data$Dam_diet_diff_PNP,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP",ylab="Effect size (Hedge's d)",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_diff_PNP, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_PNP",ylab="",xlim=c(-80,20),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_PNP[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Offspr_diet_E
symbols(data$Offspr_diet_E,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Offspr_diet_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Offspr_diet_E, na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Offspr_diet_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Offspr_diet_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_E[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Offspr_diet_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Offspr_diet_PNP_kcal
symbols(data$Offspr_diet_PNP_kcal,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Offspr_diet_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Offspr_diet_PNP_kcal[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Offspr_diet_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)

#dev.off()


####### Other continuous parameters - timing, diet differences

### Bodyweight
#pdf(file="Rplot_bubbl_BW_other.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,4),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#Bodyweight vs. TBodyweight_age_dPC
symbols(data$TBodyweight_age_dPC,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="TBodyweight_age_dPC",ylab="Effect size (Hedge's d)",main="Bodyweight - ALL",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$TBodyweight_age_dPC, na.action=na.omit), lwd=2)
symbols(data$TBodyweight_age_dPC[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="TBodyweight_age_dPC",ylab="",main="Bodyweight - F",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$TBodyweight_age_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$TBodyweight_age_dPC[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="TBodyweight_age_dPC",ylab="",main="Bodyweight - M",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$TBodyweight_age_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$TBodyweight_age_dPC[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="TBodyweight_age_dPC",ylab="",main="Bodyweight - FM",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$TBodyweight_age_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_start_exp_dPC
symbols(data$Dam_diet_start_exp_dPC,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_start_exp_dPC, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_end_exp_dPC
symbols(data$Dam_diet_end_exp_dPC,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_end_exp_dPC, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_duration
symbols(data$Dam_diet_duration,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_duration",ylab="Effect size (Hedge's d)",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_duration, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_duration[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_duration[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_duration[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_diff_E
symbols(data$Dam_diet_diff_E,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_diff_PNP_kcal
symbols(data$Dam_diet_diff_PNP_kcal,data$Bodyweight_Hd,circles=1/data$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd ~ data$Dam_diet_diff_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="F"],data$Bodyweight_Hd[data$Offspr_sex=="F"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="M"],data$Bodyweight_Hd[data$Offspr_sex=="M"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="FM"],data$Bodyweight_Hd[data$Offspr_sex=="FM"],circles=1/data$Bodyweight_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$Bodyweight_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)

#dev.off()


### RelIntake2
#pdf(file="Rplot_bubbl_RI2_other.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,4),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#RelIntake2 vs. Intake_age_dPC
symbols(data$Intake_age_dPC,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Intake_age_dPC",ylab="Effect size (Hedge's d)",main="RelIntake2 - ALL",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Intake_age_dPC, na.action=na.omit), lwd=2)
symbols(data$Intake_age_dPC[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Intake_age_dPC",ylab="",main="RelIntake2 - F",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Intake_age_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Intake_age_dPC[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Intake_age_dPC",ylab="",main="RelIntake2 - M",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Intake_age_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Intake_age_dPC[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Intake_age_dPC",ylab="",main="RelIntake2 - FM",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Intake_age_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_start_exp_dPC
symbols(data$Dam_diet_start_exp_dPC,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_start_exp_dPC, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_start_exp_dPC[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_start_exp_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_end_exp_dPC
symbols(data$Dam_diet_end_exp_dPC,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_end_exp_dPC, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_end_exp_dPC[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_end_exp_dPC[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_duration
symbols(data$Dam_diet_duration,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_duration",ylab="Effect size (Hedge's d)",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_duration, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_duration[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_duration[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_duration[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_duration[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_diff_E
symbols(data$Dam_diet_diff_E,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_E[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_E[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_diff_PNP_kcal
symbols(data$Dam_diet_diff_PNP_kcal,data$RelIntake2_Hd,circles=1/data$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd ~ data$Dam_diet_diff_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="F"],data$RelIntake2_Hd[data$Offspr_sex=="F"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="F"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="M"],data$RelIntake2_Hd[data$Offspr_sex=="M"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="M"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="M"], na.action=na.omit), lwd=2)
symbols(data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="FM"],data$RelIntake2_Hd[data$Offspr_sex=="FM"],circles=1/data$RelIntake2_Hd_Var[data$Offspr_sex=="FM"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data$RelIntake2_Hd[data$Offspr_sex=="FM"] ~ data$Dam_diet_diff_PNP_kcal[data$Offspr_sex=="FM"], na.action=na.omit), lwd=2)

#dev.off()
```

Same plots for the chow (no-choice) dam diet subset.

```{r bubble plots by sex no-choice subset, echo=FALSE, eval=TRUE, include=FALSE}
#######Dam and offspring diet parameters - no-choice data subset

### Bodyweight
#pdf(file="Rplot_bubbl_BW_NCsub.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#Bodyweight vs. Dam_diet_exp_E
symbols(data_chow$Dam_diet_exp_E,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",main="Bodyweight - ALL",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="Bodyweight - F",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="Bodyweight - M",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_exp_PNP_kcal
symbols(data_chow$Dam_diet_exp_PNP_kcal,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_contr_E
symbols(data_chow$Dam_diet_contr_E,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_contr_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_contr_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_contr_PNP_kcal
symbols(data_chow$Dam_diet_contr_PNP_kcal,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_contr_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Offspr_diet_E
symbols(data_chow$Offspr_diet_E,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Offspr_diet_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Offspr_diet_E, na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_E[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Offspr_diet_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_E[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Offspr_diet_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Offspr_diet_PNP_kcal
symbols(data_chow$Offspr_diet_PNP_kcal,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Offspr_diet_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)

#dev.off()


### RelIntake2
#pdf(file="Rplot_bubbl_RI2_NCsub.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#RelIntake2 vs. Dam_diet_exp_E
symbols(data_chow$Dam_diet_exp_E,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",main="RelIntake2 - ALL",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="RelIntake2 - F",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_E",ylab="",main="RelIntake2 - M",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_exp_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_exp_PNP_kcal
symbols(data_chow$Dam_diet_exp_PNP_kcal,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_exp_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_contr_E
symbols(data_chow$Dam_diet_contr_E,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_contr_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_contr_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_contr_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_contr_PNP_kcal
symbols(data_chow$Dam_diet_contr_PNP_kcal,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_contr_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_contr_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_contr_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Offspr_diet_E
symbols(data_chow$Offspr_diet_E,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Offspr_diet_E",ylab="Effect size (Hedge's d)",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Offspr_diet_E, na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_E[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Offspr_diet_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_E[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_E",ylab="",xlim=c(2,7),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Offspr_diet_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Offspr_diet_PNP_kcal
symbols(data_chow$Offspr_diet_PNP_kcal,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Offspr_diet_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Offspr_diet_PNP_kcal",ylab="",xlim=c(0,0.4),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Offspr_diet_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)

#dev.off()


####### Other continuous parameters - timing, diet differences

### Bodyweight
#pdf(file="Rplot_bubbl_BW_other_NCsub.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#Bodyweight vs. TBodyweight_age_dPC
symbols(data_chow$TBodyweight_age_dPC,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="TBodyweight_age_dPC",ylab="Effect size (Hedge's d)",main="Bodyweight - ALL",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$TBodyweight_age_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$TBodyweight_age_dPC[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="TBodyweight_age_dPC",ylab="",main="Bodyweight - F",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$TBodyweight_age_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$TBodyweight_age_dPC[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="TBodyweight_age_dPC",ylab="",main="Bodyweight - M",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$TBodyweight_age_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_start_exp_dPC
symbols(data_chow$Dam_diet_start_exp_dPC,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_start_exp_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_end_exp_dPC
symbols(data_chow$Dam_diet_end_exp_dPC,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_end_exp_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_duration
symbols(data_chow$Dam_diet_duration,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_duration",ylab="Effect size (Hedge's d)",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_duration, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_duration[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_duration[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_duration[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_duration[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_diff_E
symbols(data_chow$Dam_diet_diff_E,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#Bodyweight vs. Dam_diet_diff_PNP_kcal
symbols(data_chow$Dam_diet_diff_PNP_kcal,data_chow$Bodyweight_Hd,circles=1/data_chow$Bodyweight_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd ~ data_chow$Dam_diet_diff_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$Bodyweight_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$Bodyweight_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)

#dev.off()

### RelIntake2
#pdf(file="Rplot_bubbl_RI2_other_NCsub.pdf",width=7,height=8,family="sans",pointsize=10)

par(mfrow=c(6,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#RelIntake2 vs. Intake_age_dPC
symbols(data_chow$Intake_age_dPC,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Intake_age_dPC",ylab="Effect size (Hedge's d)",main="RelIntake2 - ALL",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Intake_age_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$Intake_age_dPC[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Intake_age_dPC",ylab="",main="RelIntake2 - F",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Intake_age_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Intake_age_dPC[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Intake_age_dPC",ylab="",main="RelIntake2 - M",xlim=c(0,400),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Intake_age_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_start_exp_dPC
symbols(data_chow$Dam_diet_start_exp_dPC,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_start_exp_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_start_exp_dPC",ylab="",xlim=c(-60,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_start_exp_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_end_exp_dPC
symbols(data_chow$Dam_diet_end_exp_dPC,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="Effect size (Hedge's d)",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_end_exp_dPC, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_end_exp_dPC",ylab="",xlim=c(20,50),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_end_exp_dPC[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_duration
symbols(data_chow$Dam_diet_duration,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_duration",ylab="Effect size (Hedge's d)",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_duration, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_duration[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_duration[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_duration[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_duration",ylab="",xlim=c(10,110),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_duration[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_diff_E
symbols(data_chow$Dam_diet_diff_E,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_E",ylab="Effect size (Hedge's d)",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_diff_E, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_E",ylab="",xlim=c(0,100),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_diff_E[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)
#RelIntake2 vs. Dam_diet_diff_PNP_kcal
symbols(data_chow$Dam_diet_diff_PNP_kcal,data_chow$RelIntake2_Hd,circles=1/data_chow$RelIntake2_Hd_Var,inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="Effect size (Hedge's d)",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd ~ data_chow$Dam_diet_diff_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="F"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="F"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="F"] ~ data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="F"], na.action=na.omit), lwd=2)
symbols(data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="M"],data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"],circles=1/data_chow$RelIntake2_Hd_Var[data_chow$Offspr_sex=="M"],inches=0.1,xlab="Dam_diet_diff_PNP_kcal",ylab="",xlim=c(-100,10),ylim=c(-4,6))
abline(lm(data_chow$RelIntake2_Hd[data_chow$Offspr_sex=="M"] ~ data_chow$Dam_diet_diff_PNP_kcal[data_chow$Offspr_sex=="M"], na.action=na.omit), lwd=2)

#dev.off()

```

Metafor analyses - Body weight data - Full data set
---------------------------------------------------

Controlling for the fact that some the effect sizes are affected by shared control groups. 
(using adjasted vaues saved in Hd_var_combined and variance-covariance matrices for body weight and allometric intake outcomes).


```{r metafor BW full, echo=FALSE, eval=TRUE, include=FALSE}
#From which studies the most influential data points (with smallest Hd_Var) come from?
#which(1/data$Bodyweight_Hd_Var > 10) #sample sizes 20-30 per group, but nothing dramatic
#data[which(1/data$Bodyweight_Hd_Var > 10), ]

library(metafor)

### BW
# Without moderators and without corrections for shared control groups
resBW <- rma(yi=Bodyweight_Hd, vi=Bodyweight_Hd_Var, data=data, measure="GEN", method="REML")
summary(resBW)

# Random-effect only model with study id (all below are with corrections for shared control groups)
#resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Exp_ID), data=data, method="REML", struct="UN")
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID), data=data, method="REML", struct="UN")
summary(resBW)

# Random-effect only model with study and strain (null model)
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID,~1|Strain), data=data, method="REML", struct="UN") 
summary(resBW) #overall small positive diff

# Strain model
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID), mods = ~factor(Strain)-1, data=data, method="REML", struct="UN")
summary(resBW) #Sprague-Dawley and C57BL/6 - positive large ES

# Add Sex as factor
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex)-1, data=data, method="REML", struct="UN")
summary(resBW) #largest effect in males

resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex), data=data, method="REML", struct="UN")
summary(resBW) #largest effect in males signif diff. from females

#resBW_M <- rma(yi=Bodyweight_Hd, vi=Bodyweight_matrix, data=data, measure="GEN", method="REML",subset=(Offspr_sex=="M"))#single-sex subset (nneds new VC matrix!)

#choice vs. chow diets only
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Dam_choice_diet), data=data, method="REML", struct="UN")
summary(resBW) #no diff if choice or chow dam diet


# Mixed-effect models with strain and study as random factors 
#(full models - in all males have the positive effect and it depends on the treatment end date (whether it goes into lactation))
#offspring diet type as categorical predictor
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + factor(Offspr_diet_OB) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end - 1, data=data, method="REML") 
summary(resBW)

#sex, dam diet type, offspring age and diet composition - BEST full model nr 1?
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML")
summary(resBW)

#with lactation period as categorical factor  - BEST full model nr 2?
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + Z_TBodyweight_age_dPC + factor(Dam_diet_lactation_incl) + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML") 
summary(resBW)

#with dam and offspring diet-related moderators - do not use as dam diet composition is not reliable
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) +  Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_diff_E + Z_Dam_diet_diff_PNP_kcal + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML") #adding diet properties (when available) shows that PNP matters, but diet info often missing or unreliable! - use no-choice (NC) subset next to explore diet properties effects!
summary(resBW)
```

Metafor analyses - Body weight data - chow (no-choice) diet  data subset   
-----------------------------------------------------------

```{r metafor BW chow, echo=FALSE, eval=TRUE, include=FALSE}
## No-choice subset

#exclude single FM datapoint!
data_chow68 <- subset(data_chow, Offspr_sex!="FM") #68 ES - for analyses with offspring sex moderator!
  
# Without moderators and without corrections for shared control groups
resBW <- rma(yi=Bodyweight_Hd, vi=Bodyweight_Hd_Var, data=data_chow68, measure="GEN", method="REML")
summary(resBW)

# Random-effect only model with study id (all below are with corrections for shared control groups)
#resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix, random=list(~1|Exp_ID), data=data_chow68, method="REML", struct="UN")
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID), data=data_chow68, method="REML", struct="UN")
summary(resBW)

# Random-effect only model with study and strain (null model)
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID,~1|Strain), data=data_chow68, method="REML", struct="UN") 
summary(resBW) #overall small positive diff

# Strain model
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID), mods = ~factor(Strain)-1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #Sprague-Dawley and C57BL/6 - positive large ES

# Add Sex as factor
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex)-1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #largest effect in males

resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex), data=data_chow68, method="REML", struct="UN")
summary(resBW) #largest effect in males signif diff. from females


## Mixed-effect models with strain and study as random factors 
# min
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end - 1, data=data_chow68, method="REML", struct="UN") # excluding choice studies does not change the pattern of full model - M and diet_end signif. positive
summary(resBW)

#max
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC  + Z_Dam_diet_start + Z_Dam_diet_end +  Z_Dam_diet_exp_E + Z_Dam_diet_exp_PNP_kcal +  Z_Dam_diet_contr_E + Z_Dam_diet_contr_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #diet end time ans exp PNP positive signif.

#reduced - use?
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_exp_E + Z_Dam_diet_exp_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #diet end time and exp diet PNP positive signif. - Note: PNP and E are significantly negatively correlated! (the higher energy - more fat - less protein)

#cor.test(data_chow68$Z_Dam_diet_exp_E, data_chow68$Z_Dam_diet_exp_PNP_kcal)
#par(mfrow=c(1,2),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#plot(data_chow68$Z_Dam_diet_exp_E, data_chow68$Z_Dam_diet_exp_PNP_kcal)
#abline(lm(data_chow68$Z_Dam_diet_exp_E ~ data_chow68$Z_Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2) X negative corr. between E and PNP - the higher E the less protein in the diet (more fat)!

#reduced - without E
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_exp_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #diet end time ans exp diet PNP positive signif.

#diff model
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_diff_E + Dam_diet_diff_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #same effect, but eaker, so its not the difference but avsolute values that matter?

#contr/exp model
resBW <- rma.mv(yi=Bodyweight_Hd, V=Bodyweight_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + scale(Dam_diet_contr_E) + scale(Dam_diet_contr_PNP_kcal) + scale(Dam_diet_exp_E) + scale(Dam_diet_exp_PNP_kcal) - 1, data=data_chow68, method="REML", struct="UN")
summary(resBW) #control diet not inportant but exp diett PNP is!

#raw data bubble plots
par(mfrow=c(1,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
symbols(data_chow68$Dam_diet_exp_PNP_kcal,data_chow68$Bodyweight_Hd,circles=1/data_chow68$Bodyweight_Hd_Var_combined,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$Bodyweight_Hd ~ data_chow68$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow68$Dam_diet_exp_E,data_chow68$Bodyweight_Hd,circles=1/data_chow68$Bodyweight_Hd_Var_combined,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$Bodyweight_Hd ~ data_chow68$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data_chow68$Dam_diet_end,data_chow68$Bodyweight_Hd,circles=1/data_chow68$Bodyweight_Hd_Var_combined,inches=0.1,xlab="Dam_diet_end",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$Bodyweight_Hd ~ data_chow68$Dam_diet_end, na.action=na.omit), lwd=2)


#library(lattice)
#par(mfrow=c(1,1),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#xyplot(data_chow68$Bodyweight_Hd ~ data_chow68$Dam_diet_exp_E, type=c("smooth", "p"))

confint(resBW)
funnel(resBW) #very bad funnel plot!
forest(resBW)
coef(resBW)
vcov(resBW)
```


Metafor analyses - Allometric food intake data (RI2) - Full data set
--------------------------------------------------------------------

Controlling for the fact that some the effect sizes are affected by shared control groups. 
(using adjasted vaues saved in Hd_var_combined and variance-covariance matrices for body weight and allometric intake outcomes).


```{r metafor RI2 full, echo=FALSE, eval=TRUE, include=FALSE}
#From which studies the most influential data points (with smallest Hd_Var) come from?
#which(1/data$RelIntake2_Hd_Var > 10) #sample sizes 20-30 per group, but nothing dramatic
#data[which(1/data$RelIntake2_Hd_Var > 10), ]

library(metafor)

### RI
# Without moderators and without corrections for shared control groups
resRI2 <- rma(yi=RelIntake2_Hd, vi=RelIntake2_Hd_Var, data=data, measure="GEN", method="REML")
summary(resRI2)

forest(resRI2, cex=0.5)

# Random-effect only model with study id (all below are with corrections for shared control groups)
#resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Exp_ID), data=data, method="REML", struct="UN")
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID), data=data, method="REML", struct="UN")
summary(resRI2)

# Random-effect only model with study and strain (null model)
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID,~1|Strain), data=data, method="REML", struct="UN") 
summary(resRI2) #overall small positive diff

# Strain model
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID), mods = ~factor(Strain)-1, data=data, method="REML", struct="UN")
summary(resRI2) #NMRI and C57BL/6 - positive large ES

# Add Sex as factor
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex)-1, data=data, method="REML", struct="UN")
summary(resRI2) #no sex effect

resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex), data=data, method="REML", struct="UN")
summary(resRI2) #no sex effect

#resRI2_M <- rma(yi=RelIntake2_Hd, vi=RelIntake2_matrix, data=data, measure="GEN", method="REML",subset=(Offspr_sex=="M"))#single-sex subset (needs new VC matrix!)

#choice vs. chow diets only
resRI2_M <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Dam_choice_diet), data=data, method="REML", struct="UN")
summary(resRI2_M) #no diff if choice or chow dam diet

# Mixed-effect models with strain and study as random factors 
#(full models - in all males have the positive effect and it depends on the treatment end date (whether it goes into lactation))
#offspring diet type as categorical predictor
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + factor(Offspr_diet_OB) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end - 1, data=data, method="REML") 
summary(resRI2)

#sex, dam diet type, offspring age and diet composition - BEST full model nr 1?
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML")
summary(resRI2)

#with lactation period as categorical factor  - BEST full model nr 1?
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) + Z_TBodyweight_age_dPC + factor(Dam_diet_lactation_incl) + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML") 
summary(resRI2)

#with dam and offspring diet-related moderators - do not use as dam diet composition is not reliable
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + factor(Dam_choice_diet) +  Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_diff_E + Z_Dam_diet_diff_PNP_kcal + Z_Offspr_diet_E + Z_Offspr_diet_PNP_kcal - 1, data=data, method="REML") #adding diet properties (when available) shows that PNP matters, but diet info often missing or unreliable! - use no-choice (NC) subset next to explore diet properties effects!
summary(resRI2)

funnel(resRI2) #good funnel!
```

Metafor analyses - Allometric food intake data (RI2) - no-choice data subset  
----------------------------------------------------------------------------

```{r metafor RI2 chow, echo=FALSE, eval=TRUE, include=FALSE}
## No-choice subset
#exclude single FM datapoint!
data_chow68 <- subset(data_chow, Offspr_sex!="FM") #68 ES - for analyses with offspring sex moderator!
  
# Without moderators and without corrections for shared control groups
resRI2 <- rma(yi=RelIntake2_Hd, vi=RelIntake2_Hd_Var, data=data_chow68, measure="GEN", method="REML")
summary(resRI2)

# Random-effect only model with study id (all below are with corrections for shared control groups)
#resRI2 <- rma.mv(yi=RelIntake2_Hd, V=Bodyweight_matrix, random=list(~1|Exp_ID), data=data_chow68, method="REML", struct="UN")
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID), data=data_chow68, method="REML", struct="UN")
summary(resRI2)

# Random-effect only model with study and strain (null model)
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID,~1|Strain), data=data_chow68, method="REML", struct="UN") 
summary(resRI2) #nothing

# Strain model
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID), mods = ~factor(Strain)-1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #2 mice strain - positive large ES

# Add Sex as factor
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex)-1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #nothing
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~factor(Offspr_sex), data=data_chow68, method="REML", struct="UN")
summary(resRI2) #nothing

## Mixed-effect models with strain and study as random factors 
# min
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end - 1, data=data_chow68, method="REML", struct="UN") # excluding choice studies does not change the pattern of full model - M and diet_end signif. positive
summary(resRI2) #nothing

#max
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC  + Z_Dam_diet_start + Z_Dam_diet_end +  Z_Dam_diet_exp_E + Z_Dam_diet_exp_PNP_kcal +  Z_Dam_diet_contr_E + Z_Dam_diet_contr_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #nothing

#reduced - use?
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_exp_E + Z_Dam_diet_exp_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #nothing

#cor.test(data_chow68$Z_Dam_diet_exp_E, data_chow68$Z_Dam_diet_exp_PNP_kcal)
#par(mfrow=c(1,2),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#plot(data_chow68$Z_Dam_diet_exp_E, data_chow68$Z_Dam_diet_exp_PNP_kcal)
#abline(lm(data_chow68$Z_Dam_diet_exp_E ~ data_chow68$Z_Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2) X negative corr. between E and PNP - the higher E the less protein in the diet (more fat)!

#reduced - without E
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + Z_Dam_diet_exp_PNP_kcal - 1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #males positive signif.

#contr/exp model
resRI2 <- rma.mv(yi=RelIntake2_Hd, V=RelIntake2_matrix_NC, random=list(~1|Study_ID, ~1|Strain), mods=~ factor(Offspr_sex) + Z_TBodyweight_age_dPC + Z_Dam_diet_start + Z_Dam_diet_end + scale(Dam_diet_contr_E) + scale(Dam_diet_contr_PNP_kcal) + scale(Dam_diet_exp_E) + scale(Dam_diet_exp_PNP_kcal) - 1, data=data_chow68, method="REML", struct="UN")
summary(resRI2) #males pos. signif

#raw data bubble plots
par(mfrow=c(1,3),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
symbols(data_chow68$Dam_diet_exp_PNP_kcal,data_chow68$RelIntake2_Hd,circles=1/data_chow68$RelIntake2_Hd_Var_combined,inches=0.1,xlab="Dam_diet_exp_PNP_kcal",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$RelIntake2_Hd ~ data_chow68$Dam_diet_exp_PNP_kcal, na.action=na.omit), lwd=2)
symbols(data_chow68$Dam_diet_exp_E,data_chow68$RelIntake2_Hd,circles=1/data_chow68$RelIntake2_Hd_Var_combined,inches=0.1,xlab="Dam_diet_exp_E",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$RelIntake2_Hd ~ data_chow68$Dam_diet_exp_E, na.action=na.omit), lwd=2)
symbols(data_chow68$Dam_diet_end,data_chow68$RelIntake2_Hd,circles=1/data_chow68$RelIntake2_Hd_Var_combined,inches=0.1,xlab="Dam_diet_end",ylab="Effect size (Hedge's d)",ylim=c(-4,6))
abline(lm(data_chow68$RelIntake2_Hd ~ data_chow68$Dam_diet_end, na.action=na.omit), lwd=2)


#library(lattice)
#par(mfrow=c(1,1),mar=c(4,4,1,1.75),oma=c(0,1,0,0))
#xyplot(data_chow68$RelIntake2_Hd ~ data_chow68$Dam_diet_exp_E, type=c("smooth", "p"))

confint(resRI2)
funnel(resRI2) #very nice funnel plot!
forest(resRI2)
coef(resRI2)
vcov(resRI2)
```

TO DO NEXT
----------
*use separate script for the bayesian (MCMCglmm) models, make tables
*Publication bias tests
*final figures for the MS: funnel, forest, bubble plots