Data Science Vocabulary

Data

Contextualized information

Data Science

Using data to solve problems

OSEMN “Awesome”

Obtain, Scrub, Explore, Model, iNterpret

Structured Data

Data which has some underlying organization (should be influential to the data)

Data Ethics

Moral principles

Numerical

Data that makes sense to have math applied

Categorical

Data that does not make sense to have math applied

Discrete

A type of numerical data WITH gaps

Continuous

A type of numerical data WITHOUT gaps

Nominal

A type of categorical data WITHOUT ordering (favorite color, names)

Ordinal

A type of categorical data WITH ordering (letter grades)

CSV

Comma separated values; a file format for storing tabular data

Descriptive questions

Regarding the description of past events

Predictive questions

Regarding the description of future or new events

Causal questions

Descriptions regarding potential future given and intervention

Correlation

When 2 variables are related with one another

Counterfactual

What is true under different circumstances

Observational studies

Studies that use data which was not designed for the study

Confounding factors

An unaccounted variable that affects both the treatment and control group

Randomized control trials

A type of experiment used to establish causality

Python modules

Organized collections of functions

Data type integer

A representation of integer data

Data type float

A representation of decimal data

Data type bool

A representation of true/false data

Data type string “str'“

A representation of text data

List

An ordered sequence of variables of any type of data

Dictionary

A way to represent multiple key-value pairs

Numpy array

Like a list but specifically for numbers and computation

Panda Series

A way to represent a 1D sequence of values used in pandas module

Pandas data frame

A way to represent a table of data used in pandas module

Functions

A block of code which does something

Variable scope

The portion of your data code which the variable name is valid

p th percentile

Smallest value that has at least p% of the data below it

Median

50th percentile; the middle value

Box plot

Uses percentiles to plot 1 numerical variable

Bar plot (for 1 variable)

Uses bars to plot categorical variables

Bar plot (for 2 variables)

Uses bars to plot categorical vs numerical

Histogram

Bins numerical data then makes a bar plot

Scatter plot

Uses points/dots to plot 2 numerical variables

Line plot

Plots 2 numerical variables with connected neighboring point