# DIBSI Tigers Day 3

- **Day 1-2 hackmd:** https://hackmd.io/KYExEMA4CMAYFYC0BGA7NAzIgLMgxrItHuOIusNhqhpOMrAGZA==?both

***

### Link to data carpentry lesson: http://www.datacarpentry.org/R-ecology-lesson/01-intro-to-r.html

Comic on the `sudo` command: https://imgs.xkcd.com/comics/sandwich.png

## R and Rstudio with Tracy Teal

R can do math!
`3+4`

R is so powerful because we can write scripts in the empty 4th panel in RStudio and it will create a `.R` file. That way we can save what we did!

- To make a new script you can click on the green plus sign in the top left corner in RStudio.

- Once you start typing in your rscript you can press `command enter` on the line to send it down the R console.

- **Don't forget to save your files!**
- **To comment in R use `#`** This means that R will not run whatever comes after the pound symbol.
- Comments are super helpful and will help you be a friend of yourself in the future! And also make friends with others that will use your code

# Current `tigersrock.R` file:

# assign output to a 
a <- 4

# output a
a

# assign 3 to b
b <- 3

# add a + b
a + b

# Spacing makes your code more readable 
c<-7
c

# define pi
pi <- 3.1415
round(pi) # should round to 3!

round(pi, 3)

# Specifically stating digits = 3, will help us as users be sure the the round function is doing *exactly* what we'd like! :) 
round(pi, digits = 3)

# To get help on the round function do:
?round

############ Challenge question ################
#What are the values after each statement in the following?

mass <- 47.5            # mass?
age  <- 122             # age?
mass <- mass * 2.0      # mass?
age  <- age - 20        # age?
mass_index <- mass/age  # mass_index?

# This will give us an error as we haven't defined what hello means 
hello

# It's a *bad* idea to make variable names the same name as R functions!  Avoid doing this for your safety and others'.
round <- 75.7
round(round)

# Download the survey data
download.file("https://ndownloader.figshare.com/files/2292169","portal_data_joined.csv")

# read the data to an object named surveys
surveys <- read.csv("portal_data_joined.csv")
surveys
View(surveys)

# show beginning of the data
head(surveys)
?help  # Go down to the examples for extra awesome helpful stuff!
head(surveys, n = 2)

# Structure of the data
str(surveys)

### Based on the output of str(surveys), can you answer the following questions?

#What is the class of the object surveys? It's a data.frame!

# How many rows and how many columns are in this object? 34786 rows and 13 columns/variables!
- **How many species have been recorded during these surveys?**  
    - 40 species (due to 40 factor levels)

str()

# just get the first row of data 
surveys[1,] #
surveys[1,4] # 4th column value in 1st row

# first 3 rows
surveys[1:3]

# Make a new dataframe with 100 rows
surveys_100 <- survyes[1:100,]

surveys$year

#### INSTALLING PACKAGES!
install.packages("dplyr")
library(dplyr)

select(surveys, species)
select(surveys, species, genus)
select(surveys, genus, species)

filter(surveys, years == 1995)

surveys_cool <- 
    surveys %>%
    filter(year == 1999) %>%
    select(year, species_id, weight)

str(surveys_cool)



surveys_not_cool <- 
    surveys %>%
    # We want to filter every year EXCEPT 1999
    filter(year != 1999) %>%
    select(year, species_id, weight)

str(surveys_cool)


surveys %>%
    group_by(sex) %>%
    summarize(mean_weight = mean(weight, na.rm = TRUE))

###Afternoon ####
install.packages("ggplot2")
library(ggplot2) #include the ggplot2 functions


surveys

#How to get the column year?
surveys[,4]
surveys[,"year"]
select(surveys, year) #dplyr

# How to get the rows of 1999 in base R?
subset(surveys, year == 1999)


# Pipes
surveys %>%
    head()
    
surveys %>%
    filter(weight < 5) %>%
    select(species_id, sex, weight) %>%
    head()
    
#Mutate
surveys %>%
    filter(weight < 5) %>%
    mutate(weight_kg = weight / 1000) %>% # Make a new column named weight in kg
    head()
    
?is.na
# filter for NA (keep only the rows where the weight column has an NA)
surveys %>%
    filter(is.na(weight)) %>%
    mutate(weight_kg = weight / 1000) %>% # Make a new column named weight in kg
    head()

# using ! inverts the selection. The following removes all that have NA (or keeps only the NON NA rows)
    surveys %>%
    filter(!is.na(weight)) %>%
    mutate(weight_kg = weight / 1000) %>% # Make a new column named weight in kg
    head()
    
    
    
# Pipes
surveys %>%                             # Select survey data
  filter(weight < 5) %>%                # filtering out samples that weigh less than 5
  # Only pick 3 columns 
  select(species_id, sex, weight) %>%   
  head()                                # Only view the top 6 rows 

# Mutate 
surveys %>%
  #filter(weight < 5) %>%
  filter(!is.na(weight)) %>%
  mutate(weight_kg = weight / 1000,      # Make a new column named weight in kg 
         day_of_analysis = "June_29") %>%  
  View()

# Dimensions of our data
dim(surveys)

# Spread command
surveys_gw <- 
  surveys %>%
  filter(!is.na(weight)) %>%
  group_by(genus, plot_id) %>%
  summarize(mean_weight = mean(weight))      # Calculate mean weight

head(surveys_gw)

install.packages("tidyr")
library(tidyr)
?spread


surveys_gw_wide <-  
  surveys_gw %>%
  spread(genus, mean_weight, fill = 0)
  
head(surveys_gw_wide)

surveys_gw_wide %>%
  cor(use = "pairwise.complete")


# Long formatted Data 
surveys_gw_long <- 
  surveys_gw_wide %>%
  gather(genus, mean_weight, -plot_id)

View(surveys_gw_long)



## TIME TO PLOT! 
surveys_complete <- 
  surveys %>%
  filter(species_id != "", 
         !is.na(weight),
         !is.na(hindfoot_length),
         sex != "")


surveys_plot <- 
  ggplot(data = surveys_complete,
       aes(x = weight, y = hindfoot_length)) +
  geom_point()

ggplot(data = surveys_complete,
       aes(x = weight, y = hindfoot_length)) +
  geom_point(aes(shape = sex), alpha = 0.1, color = "blue")


ggplot(data = surveys_complete,
       aes(x = weight, y = hindfoot_length)) +
  geom_point(aes(shape = sex, color = sex), alpha = 0.2, size = 3)


ggplot(data = surveys_complete,
       aes(x = weight, y = hindfoot_length)) + 
  # color based on species ID
  geom_point(aes(color = species_id), alpha = 0.1)


ggplot(data = surveys_complete,
       aes(x = species_id, y = hindfoot_length)) + 
  geom_boxplot(aes(color = species_id)) +
  theme(axis.text.x = element_text(angle = 30)) 

yearly_counts <- 
  surveys_complete %>%
  group_by(year, species_id) %>%
  tally()

ggplot(data = yearly_counts,
       aes(x = year, y = n, group = species_id, color = species_id)) + 
  geom_line() + 
  labs(y = "Species Count", x = "Year") +
  # facet_wrap
  facet_wrap(~ species_id)

##  ALL IN ONE GO 

surveys_complete %>%
  group_by(year, species_id) %>%
  tally() %>%
  ggplot(aes(x = year, y = n, group = species_id, color = species_id)) + 
  geom_line() + 
  labs(y = "Species Count", x = "Year") +
  # facet_wrap
  facet_wrap(~ species_id) + 
  theme_classic()

``


- **Figshare** is an awesome website to share your data!  
    - More information about it here: https://figshare.com/



> **Good Practice** When loading data in R, it's always good to first do a `head` and then a `str` to make sure that everything checks out fine.
>