Wine Reviews Final Project

Author

Samantha Farfsing

Introduction

This analysis will focus on different aspects that go into determining the price of various wines, and what affects the overall popularity and ratings of them. After looking through several data sets and trying to decide which to use, I found the wine reviews idea very interesting. My mom previously worked in the Wine and Spirits category for Kroger for several years, so I have had an interest in this topic for a while. At Kroger, she dealt with the pricing of wine for various states, so I was very interested to see what goes into determining these prices.

The questions that I will be using to shape my analysis include:

What varieties of wine are the most popular?
What wines have the highest pricing per bottle?
What impact does location have on pricing per bottle and overall rating?

Data

The data set I selected for the first part of my analysis was downloaded from Kaggle.

For convenience, the data set can be downloaded here in order to follow along with the analysis:

https://myxavier-my.sharepoint.com/:x:/g/personal/farfsings2_xavier_edu/EYxYzNJLtOlHvQLLxUHWUAgB2moyX5L2hpyjw2FqWDopkQ?download=1

The relevant packages needed for this analysis can be seen below.

# Loading Packages
library(tidyverse)
library(rvest)
library(lubridate)
library(httr)
library(tidytext)
library(dplyr)

Data Dictionary

Wine ID - used to keep track of each different wine in data set
Country - country that the wine was made in
Description - description of wine
Designation - vineyard within the winery that grapes are from
Points - number of points WineEnthusiast rated wine (scale of 1-100)
Price - cost per bottle of wine (in USD)
Province - province or state that wine is from
Region 1 - area that wine was made in specific province or state
Region 2 - area of region 1 that wine was made (if more specific)
Taster name - name of reviewer
Taster twitter - username of reviewer’s Twitter account
Title - name of wine
Variety - type of wine
Winery - winery that sells the wine

Descriptive Analysis

What varieties of wine are the most popular?

To analyze this question, I created a visual to show the average WineEnthusiast rating for each different variety of wine seen in the data set. First, I filtered the data to show only the top and bottom varieties of wine. After this, I created bar graphs to help visualize the highest and lowest rated varieties.

Looking at the first visual, we can see the the top 3 varieties of wine include: Tinta del Pais, Terrantez, and Gelber Traminer. These varieties all have an average rating of above 85. The second visual shows the bottom varieties of wine based on rating. The bottom 3 include Chancellor, Airen, and Aidani. After finding the top and bottom varieties based on rating, I wanted to see the average pricing for each of these. Then, I wanted to investigate if these top and bottom varieties were directly correlated to the highest and lowest prices based on variety.

Based off this bar chart, 2 out of the 3 top varieties are not seen with the highest average price. In this analysis, we can see that the average price of Ramisco wine far exceeds all the others at around $500 per bottle! The second highest average, being Terrantez which was one of the highest rated varieties, is not even half of this standing at a little less than $250 per bottle. The other varieties shown all have fairly similar average prices. After seeing these results, my next question was: what does this mean for the correlation between price and rating?

Correlation Coefficient = 0.59

This scatter plot, along with the fit line and correlation coefficient show that there is a moderately strong relationship between the average price and the average rating of wine. This was surprising to me, as I thought that the higher rated a wine was, the more expensive it would be. So, I thought the correlation would be higher for this data. After discovering this, I wanted to see how the other variables in the data set impacted the price and rating of the various wines. Also, after seeing that the top rated varieties don’t have that high of average prices, I wanted to see which varieties do hold the highest average prices, as well as which wines in general have the highest.

What wines have the highest pricing per bottle?

First, I looked at what wines overall in the data set have the highest pricing. Then, I looked the the various wine varieties and looked at the pricing of these to get an idea of the highest prices being paid among these various wines. Along with this, I also looked into the lowest priced wines overall in order to see how various variables impacted the low-priced wines as well.

The first bar graph shows the top priced wines in the data set, with the highest being the Château les Ormes Sorbet 2013 Médoc with a price of $3,300. The second highest priced wines were Château Pétrus 2014 Pomerol and Domaine du Comte Liger-Belair 2010 La Romanée, both with a price of $2,500.

The second bar graph, shows the lowest prices wines in the data set. Several of these are priced at $4, including Bandit NV Chardonnay (California), Pam’s Cuties NV Unoaked Chardonnay (California), and Terrenal 2010 Cabernet Sauvignon (Yecla). These prices are extremely low compared to the first bar graph, and would not be added to my cart when looking for wine. Although they definitely fit the college budget, this low of a price would cause hesitation in my opinion. Since I had these thoughts based off these prices, I wanted to quickly look at how the ratings of these low-priced wines looked.

I was surprised to see fairly high ratings for these wines after analyzing the prices of them. Looking closer into the data however, the lowest rated wine in this data was given 80 points, so this scale doesn’t seem to be very spread out. I would be curious to look at other ratings to see how low the ratings are for the lowest priced bottles.

Going back to the bar graphs, I noticed various repeats in the region that the wine was made. This brought about my next question to investigate, which was how location impacts the pricing and rating of wine?

How does location impact pricing and rating?

The first bar graph shows the highest average prices for wines in various countries. I was surprised to see that the highest average prices of wines in this data set are found in Australia, with Portugal being a close second. I would have expected the highest to be France or Italy. I was also surprised that the US was higher than both of these as well.

The second bar graph shows the highest average rating for wines in various countries. In this graph, Spain held the highest average ratings with Portugal following very closely behind. The US was much higher on this chart, being the third highest country. France and Italy were still at the bottom, however. Overall, it seems that the same countries are producing the top rated and highest priced wines. This could be due to climate aspects or even customer demographics. I would be interested to take this analysis even further in the future and use these topics to help analyze what else has impacts on price and ratings for wine.

Descriptive Analysis Conclusion

Overall, this analysis has shown that both location and reviewer rating have impacts on the overall price of wine. Before this analysis I had expected rating to have a higher impact on price, however. I think that there are many other factors that go into this decision, and I hope to be able to further analyze this data in the future and see what else impacts this decision.

Secondary Analysis

For the second part of my analysis, I decided to perform a sentiment analysis for some of the varieties of wine found to be the top and bottom rated in the first part of my analysis. In order to do this, I scraped the CellarTracker website. This website contains the world’s largest collection of wine reviews, tasting notes, and personal stories of people trying new wines. By collecting this data, I can see how reviewers tend to explain each of these types of wine. I will also be able to see if the sentiment scores for each variety match up with the results from the first part of my analysis.

The questions that I will be using to shape my analysis include:

What positive/negative words are most used to describe each of these varieties of wine?
How does the emotive sentiment scoring differ for each variety?

Data

The data set I selected for the second part of my analysis was scraped from CellarTracker.

For convenience, the data set can be downloaded here in order to follow along with the analysis:

https://myxavier-my.sharepoint.com/:x:/g/personal/farfsings2_xavier_edu/Edho7WVKNJ5Gi-QNc5CiY7kB8xCSLG560jygqxE_gIZBsg?download=1

The relevant packages needed for this analysis can be seen below.

# Loading Packages
library(tidyverse)
library(rvest)
library(lubridate)
library(httr)
library(tidytext)
library(dplyr)

Data Dictionary

Wine Name - name of wine reviewed
Wine Location - where the wine was made
Wine Review Date - date the wine was reviewed on CellarTracker
Wine Rating - rating reviewer gave the wine (NR = not rated)
Wine Review Content - content of review (how reviewer described it)
Wine Variety - type of wine

Sentiment Analysis

What positive/negative words are most used to describe the different wine?

First, I decided to look at the most common positive and negative words used in the reviews for each of these varieties of wine.

Looking at the graph, there seem to be more positive words used overall. The most common word used in these reviews to describe wine is sweet, followed by nice. The most common negative word used is dark, with complex following close behind.

How does the emotive sentiment scoring differ for each variety?

This chart shows us that, for all the different varieties, the amount of positive words is much higher than the other categories of words. When looking at the CellarTracker website, the reviews seemed to be fairly positive overall, so this makes sense. Also, it may not be common for reviewers to talk poorly about the wine in their reviews, and maybe they don’t take time to review wines that they don’t like. The Terrantez variety has the most positive words in its reviews. This was also one of the top rated wine categories in the first part of the analysis. This variety also seems to have trust of the reviewers, so it would be interesting to look at the different wines in the category to see the different words used to describe all the wines.

Conclusion

Overall, this analysis has shown the top and bottom wines based on price and rating. Looking at both parts of my review together, it seems that reviewers have similar opinions on the varieties of wines that are the best and worst. Based off the visuals created throughout this analysis, it seems that rating and location both have a large impact on the pricing of wine. I think that there are many more variables that go into aspects seen throughout this analysis. Although I feel that I found a lot of interesting data and creating several insights, there is much more to look into and I hope to be able to use this as a beginning to a larger analysis to find more information about this topic.

--- title: "Wine Reviews Final Project" author: "Samantha Farfsing" editor: visual format: html: code-tools: TRUE # Allow the code tools option showing in the output. self-contained: TRUE # Embeds all components into a single HTML file. execute: warning: FALSE # Prevent all warnings from displaying in the output. message: FALSE # Prevent all messages from displaying in the output. --- ## Introduction This analysis will focus on different aspects that go into determining the price of various wines, and what affects the overall popularity and ratings of them. After looking through several data sets and trying to decide which to use, I found the wine reviews idea very interesting. My mom previously worked in the Wine and Spirits category for Kroger for several years, so I have had an interest in this topic for a while. At Kroger, she dealt with the pricing of wine for various states, so I was very interested to see what goes into determining these prices. The questions that I will be using to shape my analysis include: 1. What varieties of wine are the most popular? 2. What wines have the highest pricing per bottle? 3. What impact does location have on pricing per bottle and overall rating? ## Data The data set I selected for the first part of my analysis was downloaded from Kaggle. For convenience, the data set can be downloaded here in order to follow along with the analysis: <https://myxavier-my.sharepoint.com/:x:/g/personal/farfsings2_xavier_edu/EYxYzNJLtOlHvQLLxUHWUAgB2moyX5L2hpyjw2FqWDopkQ?download=1> The relevant packages needed for this analysis can be seen below. ```{r, echo=TRUE, results='hide'} # Loading Packages library(tidyverse) library(rvest) library(lubridate) library(httr) library(tidytext) library(dplyr) ``` ```{r, echo=FALSE, include=FALSE} wine_reviews <- read_csv("SF_Final_Wine_Reviews.csv") ``` ## Data Dictionary - Wine ID - used to keep track of each different wine in data set - Country - country that the wine was made in - Description - description of wine - Designation - vineyard within the winery that grapes are from - Points - number of points WineEnthusiast rated wine (scale of 1-100) - Price - cost per bottle of wine (in USD) - Province - province or state that wine is from - Region 1 - area that wine was made in specific province or state - Region 2 - area of region 1 that wine was made (if more specific) - Taster name - name of reviewer - Taster twitter - username of reviewer's Twitter account - Title - name of wine - Variety - type of wine - Winery - winery that sells the wine ## Descriptive Analysis ### What varieties of wine are the most popular? To analyze this question, I created a visual to show the average WineEnthusiast rating for each different variety of wine seen in the data set. First, I filtered the data to show only the top and bottom varieties of wine. After this, I created bar graphs to help visualize the highest and lowest rated varieties. ```{r, echo=FALSE} bottom_varieties <- wine_reviews %>% group_by(variety) %>% summarize(avg_rating = mean(points, na.rm = TRUE)) %>% top_n(10) %>% arrange(desc(avg_rating)) %>% pull(variety) ``` ```{r, echo=FALSE} top_varieties <- wine_reviews %>% group_by(variety) %>% summarize(avg_rating = mean(points, na.rm = TRUE)) %>% top_n(10) %>% arrange(desc(avg_rating)) %>% pull(variety) wine_reviews_filtered <- wine_reviews %>% filter(variety %in% top_varieties) # Plotting ggplot(wine_reviews_filtered, aes(x = reorder(variety, points, FUN = mean), y = points)) + geom_bar(stat = "summary", fun = "mean", fill = "maroon") + labs(title = "Average Rating for Top Varieties of Wine", x = "Variety", y = "Average Rating") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) bottom_varieties <- wine_reviews %>% group_by(variety) %>% summarize(avg_rating = mean(points, na.rm = TRUE)) %>% top_n(-10) %>% arrange(avg_rating) %>% pull(variety) wine_reviews_bottom <- wine_reviews %>% filter(variety %in% bottom_varieties) ggplot(wine_reviews_bottom, aes(x = reorder(variety, points, FUN = mean), y = points)) + geom_bar(stat = "summary", fun = "mean", fill = "maroon") + labs(title = "Average Rating for Bottom Varieties of Wine", x = "Variety", y = "Average Rating") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` Looking at the first visual, we can see the the top 3 varieties of wine include: Tinta del Pais, Terrantez, and Gelber Traminer. These varieties all have an average rating of above 85. The second visual shows the bottom varieties of wine based on rating. The bottom 3 include Chancellor, Airen, and Aidani. After finding the top and bottom varieties based on rating, I wanted to see the average pricing for each of these. Then, I wanted to investigate if these top and bottom varieties were directly correlated to the highest and lowest prices based on variety. ```{r, echo=FALSE} ggplot(wine_reviews_filtered, aes(x = reorder(variety, price, FUN = mean), y = price)) + geom_bar(stat = "summary", fun = "mean", fill = "maroon") + labs(title = "Average Pricing for Top Varieties of Wine", x = "Variety", y = "Average Price") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` Based off this bar chart, 2 out of the 3 top varieties are not seen with the highest average price. In this analysis, we can see that the average price of Ramisco wine far exceeds all the others at around \$500 per bottle! The second highest average, being Terrantez which was one of the highest rated varieties, is not even half of this standing at a little less than \$250 per bottle. The other varieties shown all have fairly similar average prices. After seeing these results, my next question was: what does this mean for the correlation between price and rating? ```{r, echo=FALSE} # Calculate average rating and average price for each variety variety_summary <- wine_reviews %>% group_by(variety) %>% summarize(avg_rating = mean(points, na.rm = TRUE), avg_price = mean(price, na.rm = TRUE)) # Calculate correlation coefficient correlation <- round(cor(log10(variety_summary$avg_price), variety_summary$avg_rating, use = "pairwise.complete.obs"), 2) cat("Correlation Coefficient =", correlation, "\n") # Plotting ggplot(variety_summary, aes(x = avg_price, y = avg_rating)) + geom_point(position = position_jitter(width = 0.2, height = 0.2)) + geom_smooth(method = "lm", se = FALSE, color = "maroon") + labs(title = "Correlation between Average Price and Average Rating", x = "Average Price", y = "Average Rating") + scale_x_continuous(trans = "log10") + geom_text(x = max(log10(variety_summary$avg_price)), y = max(variety_summary$avg_rating), label = paste("Correlation coefficient:", correlation), hjust = 1, vjust = 1, color = "black", size = 3) ``` This scatter plot, along with the fit line and correlation coefficient show that there is a moderately strong relationship between the average price and the average rating of wine. This was surprising to me, as I thought that the higher rated a wine was, the more expensive it would be. So, I thought the correlation would be higher for this data. After discovering this, I wanted to see how the other variables in the data set impacted the price and rating of the various wines. Also, after seeing that the top rated varieties don't have that high of average prices, I wanted to see which varieties do hold the highest average prices, as well as which wines in general have the highest. ### What wines have the highest pricing per bottle? First, I looked at what wines overall in the data set have the highest pricing. Then, I looked the the various wine varieties and looked at the pricing of these to get an idea of the highest prices being paid among these various wines. Along with this, I also looked into the lowest priced wines overall in order to see how various variables impacted the low-priced wines as well. ```{r, echo=FALSE} # Filter out the top 10 highest priced wine titles top_priced_titles <- wine_reviews %>% arrange(desc(price)) %>% distinct(title) %>% head(10) %>% pull(title) wine_titles_filtered <- wine_reviews %>% filter(title %in% top_priced_titles) # Plotting ggplot(wine_titles_filtered, aes(x = reorder(title, -price), y = price)) + geom_bar(stat = "identity", fill = "maroon") + labs(title = "Top 10 Highest Priced Wine Titles", x = "Wine Title", y = "Price") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` ```{r, echo=FALSE} # Filter out the 10 lowest priced wine titles bottom_priced_titles <- wine_reviews %>% arrange(price) %>% distinct(title) %>% head(10) %>% pull(title) wine_titles_filtered <- wine_reviews %>% filter(title %in% bottom_priced_titles) # Plotting ggplot(wine_titles_filtered, aes(x = reorder(title, price), y = price)) + geom_bar(stat = "identity", fill = "maroon") + labs(title = "Bottom 10 Lowest Priced Wine Titles", x = "Wine Title", y = "Price") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` The first bar graph shows the top priced wines in the data set, with the highest being the Château les Ormes Sorbet 2013 Médoc with a price of \$3,300. The second highest priced wines were Château Pétrus 2014 Pomerol and Domaine du Comte Liger-Belair 2010 La Romanée, both with a price of \$2,500. The second bar graph, shows the lowest prices wines in the data set. Several of these are priced at \$4, including Bandit NV Chardonnay (California), Pam's Cuties NV Unoaked Chardonnay (California), and Terrenal 2010 Cabernet Sauvignon (Yecla). These prices are [extremely]{.underline} low compared to the first bar graph, and would not be added to my cart when looking for wine. Although they definitely fit the college budget, this low of a price would cause hesitation in my opinion. Since I had these thoughts based off these prices, I wanted to quickly look at how the ratings of these low-priced wines looked. ```{r, echo=FALSE} ggplot(wine_titles_filtered, aes(x = reorder(title, price), y = points)) + geom_point(color = "maroon") + labs(title = "Ratings for Top 10 Lowest Priced Wine Titles", x = "Wine Title", y = "Rating") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` I was surprised to see fairly high ratings for these wines after analyzing the prices of them. Looking closer into the data however, the lowest rated wine in this data was given 80 points, so this scale doesn't seem to be very spread out. I would be curious to look at other ratings to see how low the ratings are for the lowest priced bottles. Going back to the bar graphs, I noticed various repeats in the region that the wine was made. This brought about my next question to investigate, which was how location impacts the pricing and rating of wine? ### How does location impact pricing and rating? ```{r, echo=FALSE} # Calculate average pricing for each country avg_price_by_country <- wine_reviews_filtered %>% group_by(country) %>% summarize(avg_price = mean(price, na.rm = TRUE)) %>% arrange(desc(avg_price)) # Arrange countries by average price in descending order # Plotting ggplot(avg_price_by_country, aes(x = reorder(country, avg_price), y = avg_price)) + geom_bar(stat = "identity", fill = "maroon") + labs(title = "Average Pricing by Country", x = "Country", y = "Average Price") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) # Calculate average rating for each country avg_rating_by_country <- wine_reviews_filtered %>% group_by(country) %>% summarize(avg_rating = mean(points, na.rm = TRUE)) %>% arrange(desc(avg_rating)) # Arrange countries by average rating in descending order # Plotting ggplot(avg_rating_by_country, aes(x = reorder(country, avg_rating), y = avg_rating)) + geom_bar(stat = "identity", fill = "maroon") + labs(title = "Average Rating by Country", x = "Country", y = "Average Rating") + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ``` The first bar graph shows the highest average prices for wines in various countries. I was surprised to see that the highest average prices of wines in this data set are found in Australia, with Portugal being a close second. I would have expected the highest to be France or Italy. I was also surprised that the US was higher than both of these as well. The second bar graph shows the highest average rating for wines in various countries. In this graph, Spain held the highest average ratings with Portugal following very closely behind. The US was much higher on this chart, being the third highest country. France and Italy were still at the bottom, however. Overall, it seems that the same countries are producing the top rated and highest priced wines. This could be due to climate aspects or even customer demographics. I would be interested to take this analysis even further in the future and use these topics to help analyze what else has impacts on price and ratings for wine. ## Descriptive Analysis Conclusion Overall, this analysis has shown that both location and reviewer rating have impacts on the overall price of wine. Before this analysis I had expected rating to have a higher impact on price, however. I think that there are many other factors that go into this decision, and I hope to be able to further analyze this data in the future and see what else impacts this decision. ## Secondary Analysis For the second part of my analysis, I decided to perform a sentiment analysis for some of the varieties of wine found to be the top and bottom rated in the first part of my analysis. In order to do this, I scraped the CellarTracker website. This website contains the world's largest collection of wine reviews, tasting notes, and personal stories of people trying new wines. By collecting this data, I can see how reviewers tend to explain each of these types of wine. I will also be able to see if the sentiment scores for each variety match up with the results from the first part of my analysis. The questions that I will be using to shape my analysis include: 1. What positive/negative words are most used to describe each of these varieties of wine? 2. How does the emotive sentiment scoring differ for each variety? ## Data The data set I selected for the second part of my analysis was scraped from CellarTracker. For convenience, the data set can be downloaded here in order to follow along with the analysis: <https://myxavier-my.sharepoint.com/:x:/g/personal/farfsings2_xavier_edu/Edho7WVKNJ5Gi-QNc5CiY7kB8xCSLG560jygqxE_gIZBsg?download=1> The relevant packages needed for this analysis can be seen below. ```{r, echo=TRUE, results='hide'} # Loading Packages library(tidyverse) library(rvest) library(lubridate) library(httr) library(tidytext) library(dplyr) ``` ```{r, echo=FALSE, include=FALSE} combined_wine_reviews <- read_csv("combined_wine_reviews.csv") ``` ## Data Dictionary - Wine Name - name of wine reviewed - Wine Location - where the wine was made - Wine Review Date - date the wine was reviewed on CellarTracker - Wine Rating - rating reviewer gave the wine (NR = not rated) - Wine Review Content - content of review (how reviewer described it) - Wine Variety - type of wine ## Sentiment Analysis ### What positive/negative words are most used to describe the different wine? First, I decided to look at the most common positive and negative words used in the reviews for each of these varieties of wine. ```{r, echo=FALSE} # Create a review ID variable based on the date combined_wine <- combined_wine_reviews %>% arrange(wine_review_date) %>% mutate(review_id = row_number()) # unnest tokes by location and remove stop words tidy_wine_reviews <- combined_wine %>% unnest_tokens(word, wine_review_content) %>% anti_join(stop_words) # Join with the Bing lexicon to get sentiment scores tidy_wine_sentiment <- tidy_wine_reviews %>% inner_join(get_sentiments("bing"), by = "word") # Aggregate sentiment scores for each word wine_word_sentiment <- tidy_wine_sentiment %>% group_by(word) %>% summarize(sentiment_score = sum(sentiment == "positive") - sum(sentiment == "negative")) # Select top words top_words <- wine_word_sentiment %>% filter(!word == 'lemon') %>% filter(!word == 'top') %>% top_n(15, abs(sentiment_score)) # Plot the data ggplot(top_words, aes(x = reorder(word, sentiment_score), y = sentiment_score, fill = sentiment_score > 0)) + geom_bar(stat = "identity") + coord_flip() + scale_fill_manual(values = c("gray", "maroon")) + labs(title = "Top Words and Sentiment Scores", subtitle = "Based on Bing lexicon", x = "Word", y = "Sentiment Score", fill = "Sentiment") + theme_minimal() ``` Looking at the graph, there seem to be more positive words used overall. The most common word used in these reviews to describe wine is sweet, followed by nice. The most common negative word used is dark, with complex following close behind. ### How does the emotive sentiment scoring differ for each variety? ```{r, echo=FALSE} nrc <- get_sentiments("nrc") tidy_wine_reviews %>% inner_join(nrc, by = "word", relationship = "many-to-many") %>% group_by(sentiment, variety) %>% summarize(n = n()) %>% ggplot(aes(x = sentiment, y = n, fill = variety)) + geom_bar(stat = "identity", position = "dodge") + labs(title = "Variety Sentiment Scores", subtitle = "Total number of emotive words scored ", y = "Total Number of Words", x = "Emotional Sentiment", fill = "Variety") ``` This chart shows us that, for all the different varieties, the amount of positive words is much higher than the other categories of words. When looking at the CellarTracker website, the reviews seemed to be fairly positive overall, so this makes sense. Also, it may not be common for reviewers to talk poorly about the wine in their reviews, and maybe they don't take time to review wines that they don't like. The Terrantez variety has the most positive words in its reviews. This was also one of the top rated wine categories in the first part of the analysis. This variety also seems to have trust of the reviewers, so it would be interesting to look at the different wines in the category to see the different words used to describe all the wines. ## Conclusion Overall, this analysis has shown the top and bottom wines based on price and rating. Looking at both parts of my review together, it seems that reviewers have similar opinions on the varieties of wines that are the best and worst. Based off the visuals created throughout this analysis, it seems that rating and location both have a large impact on the pricing of wine. I think that there are many more variables that go into aspects seen throughout this analysis. Although I feel that I found a lot of interesting data and creating several insights, there is much more to look into and I hope to be able to use this as a beginning to a larger analysis to find more information about this topic.