Reflection on R
Jessica Ayers 2023-06-27
Reflection Post
I have worked with R in many statistics classes thus far in my academic career, but being exposed to the amount of different visuals that can be created for different processes has been the coolest thing I have learned about programming in R. Making an ordinary plot is mostly simple, but creating a plot with color, groups, different attributes, etc takes more time. In this past module 8, there was one visual that I found very interesting. It shows a regression trend for different prediction values.
Using the Seoul Bike Data:
library(ggiraphExtra)
library(tidyverse)
bikeData <- read_csv("/Users/jessayers/Documents/ST 558/TOPIC 3/SeoulBikeData.csv", locale=locale(encoding="latin1"))
bikeData
## # A tibble: 8,760 × 14
## Date `Rented Bike Count` Hour `Temperature(°C)` `Humidity(%)` `Wind speed (m/s)` `Visibility (10m)`
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 01/12/2017 254 0 -5.2 37 2.2 2000
## 2 01/12/2017 204 1 -5.5 38 0.8 2000
## 3 01/12/2017 173 2 -6 39 1 2000
## 4 01/12/2017 107 3 -6.2 40 0.9 2000
## 5 01/12/2017 78 4 -6 36 2.3 2000
## 6 01/12/2017 100 5 -6.4 37 1.5 2000
## 7 01/12/2017 181 6 -6.6 35 1.3 2000
## 8 01/12/2017 460 7 -7.4 38 0.9 2000
## 9 01/12/2017 930 8 -7.6 37 1.1 2000
## 10 01/12/2017 490 9 -6.5 27 0.5 1928
## # ℹ 8,750 more rows
## # ℹ 7 more variables: `Dew point temperature(°C)` <dbl>, `Solar Radiation (MJ/m2)` <dbl>, `Rainfall(mm)` <dbl>,
## # `Snowfall (cm)` <dbl>, Seasons <chr>, Holiday <chr>, `Functioning Day` <chr>
Similar to homework 8, we can turn the rented bike counts variable into a binary 0/1 for counts of over 700.
bikeData$count <- 0
for(i in 1:nrow(bikeData)){
if(bikeData$`Rented Bike Count`[i]>=700){
bikeData$count[i] <- 1
}
else{
bikeData$count[i] <- 0
}
}
bikeData <- bikeData %>%
rename("Temp" = "Temperature(°C)")
Now we can fit a poisson model using count as the response variable
and temperature and holiday as the predictor variables.
glmFit <- glm(count ~ Temp*Holiday, data = bikeData, family = "binomial")
Now we can plot the predicted values of this regression.
ggPredict(glmFit)
