R PACKAGES FOR DATA SCIENCE

MODULE 1 ● Bundles together code, data.

Documentation, and tests
INTRODUCTION TO DATA ANALYSIS
WITH R ● Tidyverse Library: collection of
essential R packages for data science
Value derived from data depends on
Four Steps of Data Analysis
● Accuracy of Data
● Data Wrangling and Transformation
● Accessibility of data when we need it
○ Includes: dplyr and tidyr package
Data Asset eXchange (DAX)
○ Combine different functions using
● Curated free and open datasets under
pipe operator
open data licenses
● Data Import and Management
● Provides real-world data
○ Includes: readr package
● Vetted data
○ Solves problem of parsing a flat file
● Ready to use in the enterprise
(.csv file)
● Part of developer.ibm.com
● Functional Programming
● ibm.biz/data-exchange
○ Includes: purr package
○ Provides statistics for the dataset
WHY DATA ANALYSIS (calculating mean value for each
● Data is everywhere column)
● Helps us answer questions from data ● Data Visualization and Exploration
● Plays an important role in ○ Includes: ggplot2 package
○ Discovering useful information ○ Produces charts and visualization
○ Answering questions (box plots, density plots, violin pots,
○ Predicting the future or the unknown tile pots, and time series plots)
The Problem Basic Syntax of the dplyr package
● Can you predict the likelihood of a flight ● select (): select variables by their
delay? names
● Data Analysis using R libraries for ● filter (): filter observations based on
○ Data Cleaning values
○ Exploratory Analysis ● summarize (): compute summary
○ Model Development statistics
○ Model Evaluation ● arrange (): reorder the rows
● mutate (): create new variables

UNDERSTANDING THE DATA

IMPORTING & EXPORTING DATA IN R
● Dataset-Airline Performance
○ From Data Asset eXchange Importing Data
○ In (.csv) format ● Process of loading and reading data into
● Variables in the Dataset R from various resources
○ Performing statistical analysis on ● Important Factors:
selected columns from original data ○ Format of the File (.csv, .json, .xlsx,
set .hdf)
○ File Path of the Dataset
(computer/online)
Download and Extract Data Export to Different Formats in R
● Each row is one data point (observation)
● Many properties associated with each
point
● .csv Data Format: properties are
separated from each other by commas
Load the Package in E ANALYZING DATA IN R
● Install the package
Basic Insights from the Data
● Load the tidyverse library
● Understand your dat before you begin
○ Automatically loads the readr
any analysis
package
● Check:
Import .csv Files (readr package)
○ Variable Data Types
● Includes the read_csv() function
○ Data Distribution
● Tibble: used to read .csv files into a
● Identify potential issues with data
data frame
Basic Insights of a Dataset
● Pass the location of data you want to
● Known data types in tidyverse
use to the read_csv() function as a
○ Character, date, double, integer, and
filename
logical
Help Page
● glimpse() function: determines types of
● Adding question mark before function
variables in your dataset
name
● Shows number of rows and columns in
● Documentation: includes arguments for
the dataset
the function and examples of their use
● Importance of Checking Data Types
Read the Dataset from a URL
○ Potential information and type
● Define a variable that contains the URL
mismatch
path to the file
○ Compatibility with tidyverse functions
● Download the file locally using the
R doppler::summarize(), group_by()
download file() function
● Return a statistical summary of the data
○ First Argument: URL variable
○ Statistical Metrics: tells mathematical
○ Second Argument: local name for
issues (extreme outliers and large
downloaded file
deviations)
● Unzip the content using the untar()
function
● Read the data from the local file using
the read_csv() function
Print the Data in R
● HeadFunction: shows the first 6 rows of
data frame
● Tail Function: shows the bottom 6 rows
of data frame
● Export it to a new .csv file (optional)
○ Use the write_csv() function
 How to Replace Missing Values in R
MODULE 2
● Use replace_na()
DATA WRANGLING

PRE-PROCESSING DATA IN R DATA FORMATTING IN R

Data Pre-Processing Data Formatting

● Converting or mapping data from the ● Data collected from different places and
initial raw form into another format stored in different formats
● Data cleaning or data wrangling ● Bringing data into a common standard of
Simple Data Frame Operations expression to make meaningful
● Perform data frame operations along comparisons
columns, wit beach row of the column ● Coherence: in statistics, an indication of
representing a sample the quality of the information within a
single dataset
Reformat an Entire Column
DEALING WITH MISSING VALUES IN R ● -
Missing Values Incorrect Data Types
● Missing values occur when no data ● -
value is stored for a variable in an
observation
DATA NORMALIZATION IN R
● Represented as “?”, “N/A”, 0 or just a
blank cell Data Normalization
How to Deal with Missing Data ● -
● Check with the data collection source Methods of Normalizing Data
● Drop the missing values ● -
○ Drop the variable Simple Feature Scaling in R
○ Drop the data entry ● -
● Replace the missing values Min-Max in R
○ Replace it with an average (similar ● -
data points) Z-score in R
○ Replace it with zero (frequency) ● -
○ Replace it based on other functions
● Leave it as missing data BINNING IN R
How to Check Missing Values in R
Data Normalization
● Use is.na() to count the number of
● -
missing values in columns
Methods of Normalizing Data
Drop Rows
● -
● Hyphen “-” operator: complement of a
Simple Feature Scaling in R
set in R
● -
How to Drop Missing Values in R
Min-Max in R
● Use drop_na()
● -
● Specify column names that contain
Z-score in R
missing values that you want to drop
-
Simple Feature Scaling in R
● -
●