Big Idea 2 - Data

How many values can a byte represent?

256

What is the maximum decimal value a byte can represent?

255

What is the formula to calculate how many values can be represented by n bits?

2^n

To calculate the largest number of bits use the formula:

2^n - 1

Examples of abstractions

number systems, colors, characters

Abstractions allow

• generalization of a program

• This can help shrink the code by referencing one segment instead of repeating, which lowers the chance of errors

What is the base and digit range of the octal number system?

base-8 and uses digits 0 to 7

How many symbols does the hexadecimal system use, and what are they?

base-16 and uses 0–9 and A–F (where A=10 to F=15)

Pixels

digital images are a collection of pixels that each consist of binary values

RGB pixels

represented by bytes of data, or number values up to 255, each pixel would include a 3 of these values, one for red, green, and blue

Lossy compressions

• significantly reduce file size for storage and transmission ( email ) but decreases resolution

• results in no recovery to original file, some data is lost

Lossless data compression

• if data is lost, the original file can be reproduced, but file size is larger, more difficult to store

Large datasets

may be difficult to work with, and the ability to process the data depends on the capability of the user

Bias in datasets

• not eliminated by adding more data, and algorithms that use the biased data will reflect this bias

• Large sample size could magnify bias

Cleaning

removing corrupt data, removing or repairing incomplete data, removing or flagging invalid data

Filtering

different subsets are identified and extracted to make meaning of data

Classifying

make meaning of large data sets by grouping by common attributes based on the criteria of individuals processing the data

Bias

occurs when the data collected does not represent all possibilities of the pool of available options

Patterns

the analysis starts with a question and uses rules to find patterns in data. Computers can find patterns better than people in large data sets.

Correlations

may not mean elements are caused by one another

Scalability

the capacity to increase ( or decrease ) the capacity of a resources without having to go to a completely new solution - important for storing and processing large data sets

PII ( personally identifiable information )

should be kept safe online through multi-factor identification

Metadata

Data that describes data

Used to find, organize, and manage information about data and increases the effective use of data

Changes and deletions made to metadata do not change primary data

True

Which file type is lossy compressed sound?

mp3

Which type of compression do zip files use?

Lossless

Bit

1 bit (2^0)

Byte

8 bits

Kilobyte

1,024 (2^10)

Megabyte

1,000,000 (2^20)

Gigabyte

1,000,000,000 (2^30)

Terabyte

1,000,000,000,000 (2^40)

Petabyte

1,000,000,000,000,000 (2^50)

Digital images

a collection of pixels. where each pixel consists of binary numbers.

Computers need metadata to know…

the size of an image

Data compression

is useful in computing to save disk space, or to reduce the bandwidth used when sending data (eg, over the Internet).

Text compression

• lossless compression.

• The reconstruction must be identical to the original text, as very small differences can result in statements with very different meanings.

Data extraction (first step)

the process of obtaining data from a database or software such as a social media website so that it can transport it to another software (such as spreadsheets) designed to support online analytical processing.

transform (2nd step)

Clean, filter, organize, or reformat the data into a usable structure.

analyze (3rd step)

Interpret the transformed data to find trends, patterns, or insights.

graph

A diagram used to represent data. It usually is used to depict a relationship between two things.

Picture graphs

use pictures to represent values.

Bar graphs

use either vertical or horizontal bars to represent the values.

Line graphs

use lines to represent the values.

Scatter plots

represent the data with points, and then a best-fit line is drawn through some of the points.