At the end of this week, you will be able to:
for loop, data frames, etc.The use of data as evidence is crucial but, it is not something novel. If we examine a definition of the field of statistics, we can observe that is given as four subtopics:
Originally, statistics was viewed as the analysis and interpretation of information about states. And science is understood as organized knowledge in the form of testable explanations and predictions about the universe.
So, what is data science? Data science is more than just using statistics and data to answer scientific questions.
Nowadays, data science is viewed as the use of various sources of data to extract knowledge and provide insights using multiple skills including programming, math and statistics, and communication.
Venn diagram by Drew Conway provides a visualization on data science.
Data Science Venn diagram by Drew Conway
Typical examples of data science projects:
Data science helps managers, engineers, policymakers, and researchers - almost everybody - to make informed decisions based on evidence from data. Computers and technologies have empowered how much data we can store, manipulate, and analyze. To enable these functions, technologies and tools are developed to help us to be more productive and efficient when conducting data science projects.
The technologies deployed in the analytics and data science have advanced very fast and multiple open source projects exist, for example:
Often, the data science process is iterative. Some steps in the data science workflow include:
Specify the question of interest (business understanding, scientific goal, predict or estimate,β¦)
Collect data (internal, external, sampled, relevant, ethics,β¦)
Manipulate data (explore, transform, merge, filter,β¦)
Model data (machine learning, statistics, probability, fit, validate,β¦)
Communicate and interpret the results (storytelling, visualization, dashboard, reports,β¦)
Deploy and monitor models
The two programming languages we cover in this course are R and Python. These are both open source programming languages. Letβs start off with R.
A few features of R are:
R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of UNIX platforms, Windows and MacOS. There is link to download R, documentation and manuals, The R journal, books related to R, and R packages by topics
RStudio is an integrated development environment (IDE) for R and Python, with a console, syntax-highlighting editor that supports direct code execution, and tools for plotting, history, debugging and workspace management. It compiles and runs on a wide variety of UNIX platforms, Windows and MacOS. There is an open source license that you can install for free from here: download RStudio
Quarto provides an authoring framework for data science reporting. It creates dynamic content with Python, R, Julia, and Observable and high quality reports that can be shared with an audience. Quarto (.qmd) documents are fully reproducible and support dozens of static and dynamic output formats.
For TFDS, we will be using RStudio Server hosted at UWF. This is the link https://rstudio.hmcse.uwf.edu/. Login using your UWF account.
You donβt need to install R and RStudio on your computer. But, you are welcome to do so if you wish so.
ποΈ Recordings of this week provide lessons about R, RStudio, and GitHub. The following will be covered:
# assign a value to object named "x"
x = 1
# or
x <- 1
1 -> x
# Calculator
x=10^2
y=2*x
# vectors / arrays
c(1,21,50,80,45,0)[1] 1 21 50 80 45 0# characters array
c("d","4","r")[1] "d" "4" "r"# characters
"R is useful and cool"[1] "R is useful and cool"# boolean - TRUE or FALSE
45>96[1] FALSE# built-in functions
sum(1,3,5)[1] 9# a vector / array
vec1= c(1,21,50,80,45,0)
# minimun
min(vec1)[1] 0# maximum
max(vec1)[1] 80# exponential function
exp(vec1)[1] 2.718282e+00 1.318816e+09 5.184706e+21 5.540622e+34 3.493427e+19
[6] 1.000000e+00# cosine function
cos(vec1)[1] 0.5403023 -0.5477293 0.9649660 -0.1103872 0.5253220 1.0000000# sine function
sin(vec1)[1] 0.8414710 0.8366556 -0.2623749 -0.9938887 0.8509035 0.0000000# logarithm function of base e
log(vec1,0.5)[1] 0.000000 -4.392317 -5.643856 -6.321928 -5.491853 Inf# square root
sqrt(vec1)[1] 1.000000 4.582576 7.071068 8.944272 6.708204 0.000000# logarithm function of base 10
log10(10)[1] 1# logarithm function of base 2
log2(2)[1] 1# logarithm function of base 45
log(45,base = 45)[1] 1# factorial
factorial(3)[1] 6# binomial coefficient / combination
choose(10,5)[1] 252# a set of values
vec1= c(1,21,50,80,45,0)
# summation
sum(vec1)[1] 197# arithmetic mean
mean(vec1)[1] 32.83333# standard deviation
sd(vec1)[1] 31.30122# summary statistics
summary(vec1) Min. 1st Qu. Median Mean 3rd Qu. Max.
0.00 6.00 33.00 32.83 48.75 80.00 # variance
var(vec1)[1] 979.7667# quantile
quantile(vec1,0.5)50%
33 # 100 Standard normal random numbers
x=rnorm(100,mean=0,sd=1)
# histogram
hist(x)
# create your own function
myfunction=function(){
return(print("Hello there!"))
}
# if statement
lucky.number=100
if(lucky.number<=54){
print("You win!")
}else{
print("You lost!")
}[1] "You lost!"π ποΈ Recordings on Canvas will cover more details and examples! Have fun learning and coding π! Let me know how I can help!
Instructions are posted on Canvas.