tr-student-evaluation.r

{"cells":[{"source":"<a href=\"https://www.kaggle.com/code/guleli8759/tr-student-evaluation?scriptVersionId=157030165\" target=\"_blank\"><img align=\"left\" alt=\"Kaggle\" title=\"Open in Kaggle\" src=\"https://kaggle.com/static/images/open-in-kaggle.svg\"></a>","metadata":{},"cell_type":"markdown"},{"cell_type":"markdown","id":"c8d957ac","metadata":{"_cell_guid":"0afbd355-85ec-48c9-9df0-2b4b35875a75","_uuid":"e32e5b20-c950-4412-ae5a-715306055af0","jupyter":{"outputs_hidden":false},"papermill":{"duration":0.002186,"end_time":"2023-12-29T20:28:31.99634","exception":false,"start_time":"2023-12-29T20:28:31.994154","status":"completed"},"tags":[]},"source":["\r\n","####preliminary #############\r\n","##loading  and reading packages  if its required \r\n","\r\n","if(!require(ClusterR)){install.packages(\"rstatix\")}\r\n","if(!require(cluster)){install.packages(\"cluster\")}\r\n","if(!require(factoextra)){install.packages(\"factoextra\")}\r\n","if(!require(flexclust)){install.packages(\"flexclust\")}\r\n","if(!require(fpc)){install.packages(\"fpc\")}\r\n","if(!require(ClusterR)){install.packages(\"ClusterR\")}\r\n","if(!require(ClusterR)){install.packages(\"ggpubr\")}\r\n","install.packages(\"ggpubr\", repos = \"http://cran.us.r-project.org\")\r\n","\r\n","library(cluster)\r\n","library(factoextra)\r\n","library(flexclust)\r\n","library(fpc)\r\n","library(ClusterR)\r\n","library(rstatix)\r\n","library(ggpubr)\r\n","\r\n","install.packages(\"reshape\")\r\n","library(dplyr)\r\n","library(ggplot2)\r\n","library(tidyr)\r\n","library(reshape)\r\n","\r\n","##Data Loading \r\n","library(readr)\r\n","trstudent <- read_csv(\"turkiye-student-evaluation_R_Specific.csv\")\r\n","\r\n","#view and Check first and last 6 obs of dataset to be sure that it readed Clearly \r\n","head(trstudent)\r\n","tail(trstudent)\r\n","\r\n","##Empty value controls\r\n","trstudent[!complete.cases(trstudent),]\r\n","\r\n","attach(trstudent)\r\n","#####EDA Analysis #############\r\n","\r\n","##summarize statistics and structure of dataset \r\n","str(trstudent)\r\n","summary(trstudent)\r\n","\r\n","#instance and attribute components in dataset\r\n","dim(trstudent)\r\n","\r\n","att_hist <- ggplot(trstudent, aes(x = attendance)) + geom_histogram(color = \"black\", fill = \"blue\", bins = 20) +labs(x = \"attendance\", y = \"Frequency\", title = \"Distribution of Attendance\")\r\n","easiness_hist <- ggplot(df, aes(x = Easiness)) +\r\n","  geom_histogram(color = \"black\", fill = \"green\", bins = 20) +\r\n","  stat_bin(aes(label = paste0(round((..count../nrow(df))*100), \"%\")), geom = \"text\", vjust = -0.5, color = \"white\", size = 3, bins = 20) +\r\n","  labs(x = \"Easiness\", y = \"Frequency\", title = \"Distribution of Easiness Ratings\")\r\n","\r\n","\r\n","# Plot a histogram of the scores for each question\r\n","trstudent %>%\r\n","  select(starts_with(\"Q\")) %>%\r\n","  gather() %>%\r\n","  ggplot(aes(value)) +\r\n","  geom_histogram(bins = 5) +\r\n","  facet_wrap(~key, nrow = 5)\r\n","\r\n","# Plot a boxplot of the scores for each question by course\r\n","trstudent %>%\r\n","  select(starts_with(\"Q\"), course) %>%\r\n","  gather(key = \"question\", value = \"score\", starts_with(\"Q\")) %>%\r\n","  ggplot(aes(course, score)) +  # Add the `group` aesthetic\r\n","  geom_boxplot() +\r\n","  facet_wrap(~question, nrow = 5)\r\n","\r\n","\r\n","#### Plot a heatmap of the correlation matrix\r\n","# Calculate the correlation matrix\r\n","cor_matrix <- cor(trstudent %>% select(starts_with(\"Q\")))\r\n","\r\n","# Melt the correlation matrix\r\n","cor_melted <- as.data.frame(as.table(cor_matrix))\r\n","names(cor_melted) <- c(\"Q1\", \"Q2\", \"Correlation\")\r\n","\r\n","# Create the heatmap\r\n","ggplot(cor_melted, aes(Q1, Q2, fill = Correlation)) +\r\n","  geom_tile() +\r\n","  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +\r\n","  labs(title = \"Correlation Matrix of Survey Questions\")\r\n","\r\n","# Boxplots for variables of interest\r\n","trstudent %>%\r\n","  gather(key = \"variable\", value = \"value\", 'repeat', attendance, difficulty) %>%\r\n","  ggplot(aes(x = variable, y = value)) +\r\n","  geom_boxplot() +\r\n","  labs(x = \"Variable\", y = \"Value\")\r\n","\r\n","\r\n","\r\n","## Subset the dataset to include only the evaluation questions\r\n","subset_data <- trstudent[, 7:34]\r\n","\r\n","# Scale the data to normalize the variables\r\n","scaled_data <- scale(subset_data)\r\n","\r\n","# Determine the optimal number of clusters \r\n","##using the elbow method using wcsse\r\n","\r\n","fviz_nbclust(scaled_data, FUNcluster=kmeans, method = \"wss\", k.max = 10) + theme_minimal() + ggtitle(\"The Elbow Method\")\r\n","\r\n","###using silhouette and kmeans\r\n","fviz_nbclust(scaled_data, kmeans, method=\"silhouette\")+ theme_minimal()+ ggtitle(\"The Silhouette\") # factoextra::\r\n","\r\n","###using gap statistic (*too long to calculate)\r\n","\r\n","fviz_gap_stat(clusGap(scaled_data, FUNcluster = kmeans, K.max = 24))\r\n","\r\n","\r\n","# From the elbow plot, we can see that the optimal number of clusters is 4\r\n","# Apply K-means clustering with 4 clusters\r\n","set.seed(123)\r\n","kmeans_clusters <- kmeans(scaled_data, centers=4)\r\n","\r\n","#plot(scaled_data, col = kmeans_clusters$cluster, pch=\".\", cex=3) # figure has only 2D\r\n","#points(kmeans_clusters$centers, col = 1:5, pch = 8, cex=2, lwd=2)\r\n","#fviz_cluster(list(data=scaled_data, cluster=kmeans_clusters$cluster), ellipse.type=\"norm\", geom=\"point\", stand=FALSE, palette=\"jco\", ggtheme=theme_classic()) \r\n","#factoextra::\r\n","\r\n","# View the cluster assignments for each feedback\r\n","cluster_ass <- kmeans_clusters$cluster\r\n","\r\n","centers <- kmeans_clusters$centers\r\n","# Add the cluster assignments to the original dataset\r\n","trstudent$cluster <- cluster_ass\r\n","trstudent$cluster <- cluster_ass\r\n","# Analyze the clusters by comparing the means of each variable for each cluster\r\n","cluster_means <- aggregate(subset_data, by=list(trstudent$cluster), mean)\r\n","\r\n","# View the cluster means for each variable\r\n","print(cluster_means)\r\n","\r\n","\r\n","# Compute the principal components\r\n","pca <- prcomp(eval_data, scale=TRUE)\r\n","\r\n","# Extract the first two principal components\r\n","PC1 <- pca$x[,1]\r\n","PC2 <- pca$x[,2]\r\n","help(plot)\r\n","###In this code, we first subset the dataset to only include the evaluation questions and scale the data to\r\n","#normalize the variables. We then use the prcomp() function to compute the principal components, \r\n","#with the scale=TRUE parameter indicating that the variables should be scaled. \r\n","#Finally, we extract the first two principal components (PC1 and PC2) from the result of the prcomp() function.\r\n","#Note that the principal components are linear combinations of the original variables and represent the directions\r\n","#of maximal variation in the data. The first principal component (PC1) captures the most variance in the data, \r\n","#while the second principal component (PC2) captures the second most variance. \r\n","#You can compute additional principal components by including more columns in the x[, ] indexing expression.\r\n","\r\n","\r\n","# Load the ggplot2 package\r\n","library(ggplot2)\r\n","\r\n","# Create a scatterplot of the first two principal components, colored by cluster\r\n","ggplot(trstudent, aes(x=PC1, y=PC2, color=as.factor(cluster))) +\r\n","  geom_point(size=3) +\r\n","  xlab(\"PC1\") +\r\n","  ylab(\"PC2\") +\r\n","  ggtitle(\"Clustering Results\") +\r\n","  theme_bw()\r\n","help(ggplot)\r\n","\r\n","\r\n","\r\n","# Perform hierarchical clustering\r\n","hc <- hclust(dist(survey_data), method = \"ward.D2\")\r\n","\r\n","# Plot the dendrogram\r\n","plot(hc, hang = -1, main = \"Dendrogram of Survey Data\")\r\n","\r\n","# Identify the optimal number of clusters\r\n","rect.hclust(hc, k = 5, border = \"red\")"]}],"metadata":{"kaggle":{"accelerator":"none","dataSources":[{"datasetId":3330,"sourceId":5431,"sourceType":"datasetVersion"}],"dockerImageVersionId":30618,"isGpuEnabled":false,"isInternetEnabled":true,"language":"r","sourceType":"notebook"},"kernelspec":{"display_name":"R","language":"R","name":"ir"},"language_info":{"codemirror_mode":"r","file_extension":".r","mimetype":"text/x-r-source","name":"R","pygments_lexer":"r","version":"4.0.5"},"papermill":{"default_parameters":{},"duration":3.629342,"end_time":"2023-12-29T20:28:32.122637","environment_variables":{},"exception":null,"input_path":"__notebook__.ipynb","output_path":"__notebook__.ipynb","parameters":{},"start_time":"2023-12-29T20:28:28.493295","version":"2.5.0"}},"nbformat":4,"nbformat_minor":5}