Bar plots_Kieber Lab_2023.Rmd

---
title: "Bar plots_Kieber Lab_2023"
author: "Emily J Tallerday"
date: "`r Sys.Date()`"
output: 
  powerpoint_presentation: 
    slide_level: 1
---

## Setup

This is for the R markdown file settings. You shouldn't need to change anything

```{r setup, include = F}
knitr::opts_chunk$set(echo = T , 
                      warning = F #, # including this is the ONLY way to get rid of the warnings in the flowering time plot! doing stuff inside that code chunk itself does NOT work. 
                     # message = F #,
                      # results = "hide" , 
                      # error = F ,  
                      # include = F
                      )
knitr::opts_chunk$set(tidy.opts = list(width.cutoff = 60), tidy = TRUE)
```

These are the packages we need. You will likely have to install most of these in your version of R/Rstudio

```{r, include = F}
library(rmarkdown)
#library(pandoc)
library(formatR)
packages <- c("tidyverse","gridExtra","ggpubr","viridis","ggthemes", 
              "here","gplots","cowplot","ggtext","ggsignif")
lapply(packages, require, character.only=T)
here()

library(ggokabeito) # https://ggplot2.tidyverse.org/reference/scale_colour_discrete.html
theme_set(theme_light()) # set theme
```

Set theme for ggplot

```{r, include = F}
theme_set(theme_light()) # set theme for ggplot2
```

Check location of your WD (where your data files should be):

```{r, include = F}
here()
```

## Functions

The base function for generating the bar plots with most of the settings populated. You shouldn't need to change this, but you do have to run it

```{r, include = F}
## FUNCTION to make Joe's nice bar plots (for presentations)
Bar_plot_data_by_genotype_signif_okabe <- function(data, mapping, compare, test_type) {
    ggplot(data = data, mapping = mapping) + 
    
    stat_summary(fun = mean, geom = "bar", alpha = 0.75) +
    geom_jitter(shape = 21, size = 2, alpha = 0.5, width = 0.3, height = 0.15, color = "black", fill = "black") +
    ### stats
    stat_summary(fun.data = mean_se, geom = "errorbar", color = "black", width = 0.1, size = 0.75) + # adding error bars
    geom_signif(comparisons = compare, test = test_type, 
                map_signif_level = c("***"=0.001,"**"=0.01, "*"=0.05),
                textsize = 6, color = "black", margin_top = 0.05, step_increase = 0.14, tip_length = 0.02 )
}
```

For reference

```{r, include = F}
# ## Running example
# p2 <- Bar_plot_data_by_genotype_signif_okabe(data = (double_shoot_data %>% dplyr::filter(Number_shoots > 0)), aes(x = Genotype, y = Number_shoots, color = Genotype, fill = Genotype), test_type = wilcox.test, compare = list(c("rr3,5del3,6,7,11","rr3,5,6,7,11"), c("WT", "rr3,5,6,7,11")) )
# 
# ## How to add labels, etc. ***
# p2 +
#    ylim(NA,3.5) +
#   labs(
#     title = "No. of shoots present in seedlings (7 DAG)",
#     x = "Genotype",
#     y = "No. of shoots",
#     color = "Genotype"
#   ) +
# 
# ### Setting legend names to italics while keeping the okabe coloring  
#   scale_x_discrete(labels = c("WT","<i>rr3,5del3,6,7,11</i>","<i>rr3,5,6,7,11</i>") ) +
#   theme(axis.text.x = element_markdown()) +
#   
#   scale_color_hue(labels = c("WT","<i>rr3,5del3,6,7,11</i>","<i>rr3,5,6,7,11</i>") ) + # , name = "Genotype"
#   theme(legend.text = element_markdown()) +
#   
#   scale_fill_okabe_ito(labels = c("WT","<i>rr3,5del3,6,7,11</i>","<i>rr3,5,6,7,11</i>"))
```

## Data analysis

### Load the data

I'm including some of our chlorophyll concentration data as an example -- make sure this csv file is in your current working directory (what `here()` says)

```{r, include = F}
CCI <- read.csv(file = here("Data/Opti-Sciences_CCI_full_04242023.csv") ) # read in the file
CCI <- CCI %>% na.omit() # clean it up by removing NA rows
```

NOTE: when you use your own data you'll want to format it in the "tidy" format like this csv file.

### Format data

Set names and formatting for some labels:

```{r, include = F}
## Char vec naming genotypes w MD formatting:
Genotype <- c("<i>hk3,4,5</i>", "<i>rr3,5,6,7,11</i>", "WT")
Date <- c("4/24/2023", "4/25/2023", "4/26/2023")
```

Set data classes for certain columns, as needed:

```{r, include = F}
## Manage data classes:
CCI$Rep <- as.factor(CCI$Rep)
CCI$Date <- as.factor(CCI$Date)
CCI$Genotype <- as.factor(CCI$Genotype)
CCI <- dplyr::filter(CCI, Genotype != "")
CCI$CCI <- as.numeric(CCI$CCI)
```

### Prep stats to run

Set what stats you want to run, and which things to compare. See [here](https://stackoverflow.com/questions/57567118/how-to-create-in-r-a-vector-automatically-which-contains-all-groups-to-be-compar) for details. The following will generate a list with all possible `Genotype` names paired (`2`), and this will be used to run the t.tests/equivalent later on.

```{r, include = F}
pairs <- combn(as.character(unique(CCI$Genotype)), 2, simplify = F)
```

# Making your plot

```{r, echo = F, dev = "png", dpi = 300}
Bar_plot_data_by_genotype_signif_okabe(
  data = CCI %>% 
    dplyr::filter(Date == "4/26/2023" ), # df to plot from, and I've filtered it so only data from that date are included

  ### Set axis variables and coloring:
  aes(x = Genotype, 
      y = CCI, 
      color = Genotype, 
      fill = Genotype), # color = line color, fill = bar fill color

  ### Give test info:  
  test_type = wilcox.test, # the default is a standart t-test. I've set wilcox here because we have more than 2 genotypes
  compare = pairs # This must match the list "pairs" made above
) +


  ### Adjust labels:  
  labs(
    title = "Chlorophyll content in 9-week old plants after 5 days dark treatment",
    x = "Genotype",
    y = "Chlorophyll content (CCI)",
    color = "Genotype", # in legend
    fill = "Genotype" # in legend
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust=1)) + # turns x-axis labels 45 degrees to fit


  ### Setting legend names to italics while keeping the special okabe coloring
  # just make sure these line up with what you set to color and fill at the top!
  scale_x_discrete(labels = Genotype) + # x-axis variable 
  theme(axis.text.x = element_markdown()) + # will take MD formatting you provided to do italics/bold/whatever -- here I've set it above to make the mutant names in italics
  scale_color_hue(labels = Genotype) + # color variable
  theme(legend.text = element_markdown()) +
  scale_fill_okabe_ito(labels = Genotype) # fill variable
```

You can copy and paste the plot above into another document. This document is set to make a presentation, and all plots generated will be made as png's with a DPI of 300 (`dev = "png", dpi = 300` at top of this plot chunk).

```{r}
# knit doc early to exclude everything below from going into ppt file
knitr::knit_exit()
```

## Trying to optimize stats visuals with iris dataset

```{r, include = F}
## Char vec naming genotypes w MD formatting:
Species <- c("<i>setosa</i>", "<i>versicolor</i>", "virginica")
```

```{r, echo = F, dev = "png", dpi = 300}
# pairs <- combn(as.character(unique(iris$Species)), 2, simplify = F)
# 
# Bar_plot_data_by_genotype_signif_okabe(
#   data = iris %>% 
#       ### Set axis variables and coloring:
#   aes(x = Species, 
#       y = `Petal.Length`, 
#       color = Species, 
#       fill = Species), # color = line color, fill = bar fill color
# 
#   ### Give test info:  
#   test_type = wilcox.test, # the default is a standart t-test. I've set wilcox here because we have more than 2 genotypes
#   compare = pairs # This must match the list "pairs" made above
# ) +
# 
#   ### Adjust labels:  
#   labs(
#     title = "",
#     x = "Species",
#     y = "Petal Length",
#     color = "Species", # in legend
#     fill = "Species" # in legend
#   ) +
#   theme(axis.text.x = element_text(angle = 45, hjust=1)) + # turns x-axis labels 45 degrees to fit
# 
# 
#   ### Setting legend names to italics while keeping the special okabe coloring
#   # just make sure these line up with what you set to color and fill at the top!
#   scale_x_discrete(labels = Species) + # x-axis variable 
#   theme(axis.text.x = element_markdown()) + # will take MD formatting you provided to do italics/bold/whatever -- here I've set it above to make the mutant names in italics
#   scale_color_hue(labels = Species) + # color variable
#   theme(legend.text = element_markdown()) +
#   scale_fill_okabe_ito(labels = Species) # fill variable
```

## Session info

```{r, include = F}
sessionInfo()
```