Pedaci HW2

index.qmd

@@ -1,11 +1,10 @@
 ---
 title: "Homework 2"
-author: "[Insert your name here]{style='background-color: yellow;'}"
+author: "Carson Pedaci {style='background-color: yellow;'}"
 toc: true
 title-block-banner: true
 title-block-style: default
-format: html
-# format: pdf
+format: pdf
 ---
 
 [Link to the Github repository](https://github.com/psu-stat380/hw-2)
@@ -28,7 +27,7 @@ For this assignment, we will be using the [Abalone dataset](http://archive.ics.u
 
 We will be using the following libraries:
 
-```R
+```{R}
 library(readr)
 library(tidyr)
 library(ggplot2)
@@ -51,7 +50,7 @@ EDA using `readr`, `tidyr` and `ggplot2`
 
 Load the "Abalone" dataset as a tibble called `abalone` using the URL provided below. The `abalone_col_names` variable contains a vector of the column names for this dataset (to be consistent with the R naming pattern). Make sure you read the dataset with the provided column names. 
 
-```R
+```{R}
 library(readr)
 url <- "http://archive.ics.uci.edu/ml/machine-learning-databases/abalone/abalone.data"
 
@@ -67,7 +66,8 @@ abalone_col_names <- c(
   "rings"
 )
 
-abalone <- ... # Insert your code here
+abalone <- readr::read_csv(url, col_names = abalone_col_names)
+abalone
 ```
 
 ---
@@ -76,11 +76,12 @@ abalone <- ... # Insert your code here
 
 Remove missing values and `NA`s from the dataset and store the cleaned data in a tibble called `df`. How many rows were dropped?
 
-```R
-df <- ... # Insert your code here
+```{R}
+df <- abalone[complete.cases(abalone), ]
+df
 ```
 
-
+No rows were dropped.
 
 
 ---
@@ -89,8 +90,12 @@ df <- ... # Insert your code here
 
 Plot histograms of all the quantitative variables in a **single plot** [^footnote_facet_wrap]
 
-```R
-... # Insert your code here
+```{R}
+df %>%
+  pivot_longer(cols = c('length', 'diameter', 'height', 'whole_weight', 'shucked_weight', 'viscera_weight', 'shell_weight', 'rings'), names_to = 'var', values_to = 'val')%>%
+  ggplot +
+  geom_histogram(aes(x = val)) +
+  facet_wrap(facets = 'var')
 ```
 
 
@@ -100,12 +105,16 @@ Plot histograms of all the quantitative variables in a **single plot** [^footnot
 
 Create a boxplot of `length` for each `sex` and create a violin-plot of of `diameter` for each `sex`. Are there any notable differences in the physical appearences of abalones based on your analysis here?
 
-```R
-... # Insert your code for boxplot here
+```{R}
+df %>%
+  ggplot +
+  geom_boxplot(aes(x = sex, y = length))
 ```
 
-```R
-... # Insert your code for violinplot here
+```{R}
+df %>%
+  ggplot +
+  geom_violin(aes(x = sex, y = diameter))
 ```
 
 
@@ -117,19 +126,25 @@ Create a scatter plot of `length` and `diameter`, and modify the shape and color
 
 
 
-```R
-... # Insert your code here
+```{R}
+df %>%
+  ggplot +
+  geom_point(aes(x = length, y = diameter, color = sex, size = shell_weight))
 ```
-
+As length increases, so does the shell weight. There is one abalone with a significantly greater diameter compared to other abalones of its length.
 ---
 
 ###### 1.6 (5 points)
 
 For each `sex`, create separate scatter plots of `length` and `diameter`. For each plot, also add a **linear** trendline to illustrate the relationship between the variables. Use the `facet_wrap()` function in R for this, and ensure that the plots are vertically stacked **not** horizontally. You should end up with a plot that looks like this: [^footnote_plot_facet]
 
 
-```R
-... # Insert your code here
+```{R}
+df %>%
+  ggplot +
+  geom_point(aes(x = length, y = diameter)) +
+  facet_grid(facets = 'sex', as.table = TRUE) +
+  geom_smooth(aes(x = length, y = diameter))
 ```
 
 
@@ -154,8 +169,11 @@ More advanced analyses using `dplyr`, `purrrr` and `ggplot2`
 Filter the data to only include abalone with a length of at least $0.5$ meters. Group the data by `sex` and calculate the mean of each variable for each group. Create a bar plot to visualize the mean values for each variable by `sex`.
 
 
-```R
-df %>% ... # Insert your code here
+```{R}
+df %>%
+  filter(length >= 0.5) %>%
+  group_by(sex)%>%
+  summarise(mean(length), mean(diameter), mean(height), mean(whole_weight), mean(shucked_weight), mean(viscera_weight), mean(shell_weight), mean(rings))
 ```
 
 
@@ -176,8 +194,16 @@ Implement the following in a **single command**:
 3. Use the `geom_tile()` function to create a tile plot of `num_rings` vs `sex` with the color indicating of each tile indicating the `avg_weight` value. 
 
 
-```R
-df %>% ... # Insert your code here
+```{R}
+num_df <-
+  df %>%
+  mutate('num_rings' = case_when(rings < 10 ~ 'low', rings > 20 ~ 'high', 10 <= rings | rings <= 20 ~ 'med'))
+
+num_df %>%
+  group_by(num_rings, sex) %>%
+  summarise('avg_weight' = mean(whole_weight + shucked_weight + viscera_weight + shell_weight))%>%
+  ggplot +
+  geom_tile(aes(x = sex, y = num_rings, fill = avg_weight))
 ```
 
 
@@ -189,8 +215,11 @@ df %>% ... # Insert your code here
 Make a table of the pairwise correlations between all the numeric variables rounded to 2 decimal points. Your final answer should look like this [^footnote_table]
 
 
-```R
-df %>% ... # Insert your code here
+```{R}
+R <- df %>%
+keep(is.numeric) %>%
+cor()
+R
 ```
 
 
@@ -206,8 +235,12 @@ Use the `map2()` function from the `purrr` package to create a scatter plot for
 :::
 
 
-```R
-... # Insert your code here
+```{R}
+df %>%
+  pivot_longer(cols = c('length', 'diameter', 'height', 'whole_weight', 'shucked_weight',   'viscera_weight', 'shell_weight'), names_to = 'var', values_to = 'val')%>%
+  ggplot +
+  geom_point(aes(x = rings, y = val, color = sex)) +
+  facet_grid(facets = 'var')
 ```
 
 
@@ -230,8 +263,10 @@ Linear regression using `lm`
 Perform a simple linear regression with `diameter` as the covariate and `height` as the response. Interpret the model coefficients and their significance values. 
 
 
-```R
-... # Insert your code here
+```{R}
+model <- lm(df$height ~ df$diameter)
+
+summary(model)
 ```
 
 
@@ -243,10 +278,12 @@ Perform a simple linear regression with `diameter` as the covariate and `height`
 Make a scatterplot of `height` vs `diameter` and plot the regression line in `color="red"`. You can use the base `plot()` function in R for this. Is the linear model an appropriate fit for this relationship? Explain. 
 
 
-```R
-... # Insert your code here
+```{R}
+x = df$diameter
+plot(df$diameter, df$height)
+lines(df$diameter, fitted(lm(df$height~df$diameter)), col='red')
 ```
-
+The linear model is an appropriate fit. The data present no visible patterns of a correlation other than linear.
 
 
 ---
@@ -255,8 +292,8 @@ Make a scatterplot of `height` vs `diameter` and plot the regression line in `co
 
 Suppose we have collected observations for "new" abalones with `new_diameter` values given below. What is the expected value of their `height` based on your model above? Plot these new observations along with your predictions in your plot from earlier using `color="violet"`
 
-```R
-
+```{R}
+df3 <- df %>% select(diameter, height)
 new_diameters <- c(
   0.15218946,
   0.48361548,
@@ -268,10 +305,14 @@ new_diameters <- c(
   0.92906875,
   0.94245437,
   0.01209518
-)
-
-
-... # Insert your code here. 
+) 
+
+nd_frame = data.frame(new_diameters)%>%
+  rename('diameter' = 'new_diameters')
+model <- lm(height ~ diameter, data = df3)
+predict_model <- predict(model, nd_frame)
+plot(nd_frame%>% unlist(), predict_model, col = 'violet')
+points(df3$diameter, df3$height, col = 'grey')
 ```