shmick/R
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Second attempt

Browse Source
This commit is contained in:
Matan Horovitz
2023-02-26 17:50:25 +02:00
parent a40330b3ac
commit 6bd14b5476
1 changed files with 17 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
R_Final_Tasks_Statistics.R
View File

@@ -1,32 +1,23 @@
##Final R assignment in Intro to Statistics course, fall semster. ##Final R assignment in Intro to Statistics course, fall semster.
#+Written by Matan Horovitz (207130253) and Guy Amzaleg () #+Written by Matan Horovitz (207130253) and Guy Amzaleg ()
#+We have chosen a dataset of the first 600,000 commits to the Linux Kernel Git repository - as published on Kaggle: #+We have chosen a dataset of CPU and GPU performance trends since 2000 - as published on Kaggle:
#+https://www.kaggle.com/datasets/philschmidt/linux-kernel-git-revision-history #+https://www.kaggle.com/datasets/michaelbryantds/cpu-and-gpu-product-data
#+This dataset examines commits made to the Linux kernel project over the last 12 years.
raw_kernel_commits <- read.csv("/home/shmick/linux_kernel_git_revlog.csv")
##BONUS: convert from EPOCH: as.Date(as.POSIXct(1100171890,origin = "1970-01-01"))
View(raw_kernel_commits)
##For question 1, we have chosen to examine whether the amount of changes (additions and deletions) vary over time.
##A larger amount of changes per commit is correlated with a decrease in the quality of the code - as more changes
#+are harder to track and audit, while a smaller amount of changes per commit implies stricter coding standards
#+over time.
##To examine the correlation, we made a subset of the dataset, discarding data which is irrelevant to our question,
#+and summed all changes from the same commits into an aggregate - as we are not concerned about the files changed,
#+only the total amount of changes on each individual commit.
kernel_commits <- subset(raw_kernel_commits, select= c(author_timestamp,n_additions,n_deletions)) raw_perf_data <- read.csv("/home/shmick/Downloads/chip_dataset.csv")
# Make a subset ^ of the dataset which includes ^ the EPOCH time,^ additions ^ and deletions. ##BONUS: convert from EPOCH: as.Date(as.POSIXct(1100171890,origin = "1970-01-01"))
#Rename the resulting dataset into friendlier column names View(raw_perf_data)
colnames(kernel_commits) <- c("EPOCH","Additions","Deletions") ##For question 1, we have chosen to examine which type of chip has examined the greater improvement over the years - GPU chips or CPU chips.
#Sum additions and deletion into a new column, named "Total changes" #+As chip perfomance is most directly correlated with the number of transistors, we have measured the pace of development based on pace of
kernel_commits["Total changes"] <- kernel_commits["Additions"]+kernel_commits["Deletions"] #+increasing transistor count.
View(kernel_commits) #< examine the resulting dataset CPU <- chip[chip$Type == 'CPU',]
#Unite all commits with same EPOCH to get changes PER COMMIT GPU <- chip[chip$Type == 'GPU',]
kernel_commits_sum <- aggregate(. ~ EPOCH,data=kernel_commits,FUN=sum)
CPU_Transistor_Count <- order(CPU$Transistors..million.)
View(kernel_commits_sum) GPU_Transistor_Count <- order(GPU$Transistors..million.)
#Measure the correlation between EPOCH time (from oldest to newest) and the total number of changes per commit. ##Iterate over date entries
cor(kernel_commits_sum["EPOCH"],kernel_commits_sum["Total changes"]) for (i in 1:length(CPU$Release.Date)){print(i)}
##Get date
for (i in 1:length(CPU$Release.Date)){print(CPU$Release.Date[i])}
##QUESTION 2: measure number of columns in our dataset and calculate a permutation and combination of ##QUESTION 2: measure number of columns in our dataset and calculate a permutation and combination of
#+that number, minus two, and 3. #+that number, minus two, and 3.