37,19 €
Mehr erfahren.
- Herausgeber: Packt Publishing
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
- Veröffentlichungsjahr: 2017
Get your statistics basics right before diving into the world of data science
About This Book
- No need to take a degree in statistics, read this book and get a strong statistics base for data science and real-world programs;
- Implement statistics in data science tasks such as data cleaning, mining, and analysis
- Learn all about probability, statistics, numerical computations, and more with the help of R programs
Who This Book Is For
This book is intended for those developers who are willing to enter the field of data science and are looking for concise information of statistics with the help of insightful programs and simple explanation. Some basic hands on R will be useful.
What You Will Learn
- Analyze the transition from a data developer to a data scientist mindset
- Get acquainted with the R programs and the logic used for statistical computations
- Understand mathematical concepts such as variance, standard deviation, probability, matrix calculations, and more
- Learn to implement statistics in data science tasks such as data cleaning, mining, and analysis
- Learn the statistical techniques required to perform tasks such as linear regression, regularization, model assessment, boosting, SVMs, and working with neural networks
- Get comfortable with performing various statistical computations for data science programmatically
In Detail
Data science is an ever-evolving field, which is growing in popularity at an exponential rate. Data science includes techniques and theories extracted from the fields of statistics; computer science, and, most importantly, machine learning, databases, data visualization, and so on.
This book takes you through an entire journey of statistics, from knowing very little to becoming comfortable in using various statistical methods for data science tasks. It starts off with simple statistics and then move on to statistical methods that are used in data science algorithms. The R programs for statistical computation are clearly explained along with logic. You will come across various mathematical concepts, such as variance, standard deviation, probability, matrix calculations, and more. You will learn only what is required to implement statistics in data science tasks such as data cleaning, mining, and analysis. You will learn the statistical techniques required to perform tasks such as linear regression, regularization, model assessment, boosting, SVMs, and working with neural networks.
By the end of the book, you will be comfortable with performing various statistical computations for data science programmatically.
Style and approach
Step by step comprehensive guide with real world examples
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 332
Ähnliche
BIRMINGHAM - MUMBAI
Statistics for Data Science
Copyright © 2017 Packt Publishing
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without the prior written permission of the publisher, except in the case of brief quotations embedded in critical articles or reviews.
Every effort has been made in the preparation of this book to ensure the accuracy of the information presented. However, the information contained in this book is sold without warranty, either express or implied. Neither the author, nor Packt Publishing, and its dealers and distributors will be held liable for any damages caused or alleged to be caused directly or indirectly by this book.
Packt Publishing has endeavored to provide trademark information about all of the companies and products mentioned in this book by the appropriate use of capitals. However, Packt Publishing cannot guarantee the accuracy of this information.
First published: November 2017
Production reference: 1151117
ISBN 978-1-78829-067-8
www.packtpub.com
Credits
Author
James D. Miller
Copy Editor
Tasneem Fatehi
Reviewers
James C. Mott
Project Coordinator
Manthan Patel
Commissioning Editor
Veena Pagare
Proofreader
Safis Editing
Acquisition Editor
Tushar Gupta
Indexer
Aishwarya Gangawane
Content Development Editor
Snehal Kolte
Graphics
Tania Dutta
Technical EditorSayli Nikalje
Production Coordinator
Deepika Naik
About the Author
James D. Miller, is an IBM certified expert, creative innovator and accomplished Director, Sr. Project Leader and Application/System Architect with +35 years of extensive applications and system design and development experience across multiple platforms and technologies. Experiences include introducing customers to new and sometimes disruptive technologies and platforms, integrating with IBM Watson Analytics, Cognos BI, TM1 and web architecture design, systems analysis, GUI design and testing, database modelling and systems analysis, design and development of OLAP, client/server, web and mainframe applications and systems utilizing: IBM Watson Analytics, IBM Cognos BI and TM1 (TM1 rules, TI, TM1Web and Planning Manager), Cognos Framework Manager, dynaSight-ArcPlan, ASP, DHTML, XML, IIS, MS Visual Basic and VBA, Visual Studio, PERL, SPLUNK, WebSuite, MS SQL Server, ORACLE, SYBASE Server, and so on.
Responsibilities have also included all aspects of Windows and SQL solution development and design including analysis; GUI (and website) design; data modelling; table, screen/form and script development; SQL (and remote stored procedures and triggers) development/testing; test preparation and management and training of programming staff. Other experience includes the development of Extract, Transform, and Load (ETL) infrastructure such as data transfer automation between mainframe (DB2, Lawson, Great Plains, and so on.) systems and client/server SQL server and web-based applications and integration of enterprise applications and data sources.
Mr Miller has acted as Internet Applications Development Mgr. responsible for the design, development, QA and delivery of multiple websites including online trading applications, warehouse process control and scheduling systems, administrative and control applications. Mr Miller also was responsible for the design, development and administration of a web-based financial reporting system for a 450-million-dollar organization, reporting directly to the CFO and his executive team.
He has also been responsible for managing and directing multiple resources in various management roles including project and team leader, lead developer and applications development director.
He has authored the following books published by Packt:
Mastering Predictive Analytics with R – Second Edition
Big Data Visualization
Learning IBM Watson Analytics
Implementing Splunk – Second Edition
Mastering Splunk
IBM Cognos TM1 Developer's Certification Guide
He has also authored a number of whitepapers on best practices such as Establishing a Center of Excellence and continues to post blogs on a number of relevant topics based on personal experiences and industry best practices.
He is a perpetual learner continuing to pursue experiences and certifications, currently holding the following current technical certifications:
IBM Certified Developer Cognos TM1
IBM Certified Analyst Cognos TM1
IBM Certified Administrator Cognos TM1
IBM Cognos TM1 Master 385 Certification
IBM Certified Advanced Solution Expert Cognos TM1
IBM OpenPages Developer Fundamentals C2020-001-ENU
IBM Cognos 10 BI Administrator C2020-622
IBM Cognos 10 BI Author C2090-620-ENU
IBM Cognos BI Professional C2090-180-ENU
IBM Cognos 10 BI Metadata Model Developer C2090-632
IBM Certified Solution Expert - Cognos BI
Specialties: The evaluation and introduction of innovative and disruptive technologies, cloud migration, IBM Watson Analytics, big data, data visualizations, Cognos BI and TM1 application design and development, OLAP, Visual Basic, SQL Server, forecasting and planning; international application, and development, business intelligence, project development, and delivery and process improvement.
About the Reviewer
James Mott, Ph.D, is a senior education consultant with extensive experience in teaching statistical analysis, modeling, data mining and predictive analytics. He has over 30 years of experience using SPSS products in his own research including IBM SPSS Statistics, IBM SPSS Modeler, and IBM SPSS Amos. He has also been actively teaching these products to IBM/SPSS customers for over 30 years. In addition, he is an experienced historian with expertise in the research and teaching of 20th Century United States political history and quantitative methods. His specialties are data mining, quantitative methods, statistical analysis, teaching, and consulting.
www.PacktPub.com
For support files and downloads related to your book, please visit www.PacktPub.com.
Did you know that Packt offers eBook versions of every book published, with PDF and ePub files available? You can upgrade to the eBook version at www.PacktPub.com and as a print book customer, you are entitled to a discount on the eBook copy. Get in touch with us at [email protected] for more details.
At www.PacktPub.com, you can also read a collection of free technical articles, sign up for a range of free newsletters and receive exclusive discounts and offers on Packt books and eBooks.
https://www.packtpub.com/mapt
Get the most in-demand software skills with Mapt. Mapt gives you full access to all Packt books and video courses, as well as industry-leading tools to help you plan your personal development and advance your career.
Why subscribe?
Fully searchable across every book published by Packt
Copy and paste, print, and bookmark content
On demand and accessible via a web browser
Customer Feedback
Thanks for purchasing this Packt book. At Packt, quality is at the heart of our editorial process. To help us improve, please leave us an honest review on this book's Amazon page at https://www.amazon.com/dp/1788290674. If you'd like to join our team of regular reviewers, you can email us at [email protected]. We award our regular reviewers with free eBooks and videos in exchange for their valuable feedback. Help us be relentless in improving our products!
Table of Contents
Preface
What this book covers
What you need for this book
Who this book is for
Conventions
Reader feedback
Customer support
Downloading the example code
Downloading the color images of this book
Errata
Piracy
Questions
Transitioning from Data Developer to Data Scientist
Data developer thinking
Objectives of a data developer
Querying or mining
Data quality or data cleansing
Data modeling
Issue or insights
Thought process
Developer versus scientist
New data, new source
Quality questions
Querying and mining
Performance
Financial reporting
Visualizing
Tools of the trade
Advantages of thinking like a data scientist
Developing a better approach to understanding data
Using statistical thinking during program or database designing
Adding to your personal toolbox
Increased marketability
Perpetual learning
Seeing the future
Transitioning to a data scientist
Let's move ahead
Summary
Declaring the Objectives
Key objectives of data science
Collecting data
Processing data
Exploring and visualizing data
Analyzing the data and/or applying machine learning to the data
Deciding (or planning) based upon acquired insight
Thinking like a data scientist
Bringing statistics into data science
Common terminology
Statistical population
Probability
False positives
Statistical inference
Regression
Fitting
Categorical data
Classification
Clustering
Statistical comparison
Coding
Distributions
Data mining
Decision trees
Machine learning
Munging and wrangling
Visualization
D3
Regularization
Assessment
Cross-validation
Neural networks
Boosting
Lift
Mode
Outlier
Predictive modeling
Big Data
Confidence interval
Writing
Summary
A Developer's Approach to Data Cleaning
Understanding basic data cleaning
Common data issues
Contextual data issues
Cleaning techniques
R and common data issues
Outliers
Step 1 – Profiling the data
Step 2 – Addressing the outliers
Domain expertise
Validity checking
Enhancing data
Harmonization
Standardization
Transformations
Deductive correction
Deterministic imputation
Summary
Data Mining and the Database Developer
Data mining
Common techniques
Visualization
Cluster analysis
Correlation analysis
Discriminant analysis
Factor analysis
Regression analysis
Logistic analysis
Purpose
Mining versus querying
Choosing R for data mining
Visualizations
Current smokers
Missing values
A cluster analysis
Dimensional reduction
Calculating statistical significance
Frequent patterning
Frequent item-setting
Sequence mining
Summary
Statistical Analysis for the Database Developer
Data analysis
Looking closer
Statistical analysis
Summarization
Comparing groups
Samples
Group comparison conclusions
Summarization modeling
Establishing the nature of data
Successful statistical analysis
R and statistical analysis
Summary
Database Progression to Database Regression
Introducing statistical regression
Techniques and approaches for regression
Choosing your technique
Does it fit?
Identifying opportunities for statistical regression
Summarizing data
Exploring relationships
Testing significance of differences
Project profitability
R and statistical regression
A working example
Establishing the data profile
The graphical analysis
Predicting with our linear model
Step 1: Chunking the data
Step 2: Creating the model on the training data
Step 3: Predicting the projected profit on test data
Step 4: Reviewing the model
Step 4: Accuracy and error
Summary
Regularization for Database Improvement
Statistical regularization
Various statistical regularization methods
Ridge
Lasso
Least angles
Opportunities for regularization
Collinearity
Sparse solutions
High-dimensional data
Classification
Using data to understand statistical regularization
Improving data or a data model
Simplification
Relevance
Speed
Transformation
Variation of coefficients
Casual inference
Back to regularization
Reliability
Using R for statistical regularization
Parameter Setup
Summary
Database Development and Assessment
Assessment and statistical assessment
Objectives
Baselines
Planning for assessment
Evaluation
Development versus assessment
Planning
Data assessment and data quality assurance
Categorizing quality
Relevance
Cross-validation
Preparing data
R and statistical assessment
Questions to ask
Learning curves
Example of a learning curve
Summary
Databases and Neural Networks
Ask any data scientist
Defining neural network
Nodes
Layers
Training
Solution
Understanding the concepts
Neural network models and database models
No single or main node
Not serial
No memory address to store results
R-based neural networks
References
Data prep and preprocessing
Data splitting
Model parameters
Cross-validation
R packages for ANN development
ANN
ANN2
NNET
Black boxes
A use case
Popular use cases
Character recognition
Image compression
Stock market prediction
Fraud detection
Neuroscience
Summary
Boosting your Database
Definition and purpose
Bias
Categorizing bias
Causes of bias
Bias data collection
Bias sample selection
Variance
ANOVA
Noise
Noisy data
Weak and strong learners
Weak to strong
Model bias
Training and prediction time
Complexity
Which way?
Back to boosting
How it started
AdaBoost
What you can learn from boosting (to help) your database
Using R to illustrate boosting methods
Prepping the data
Training
Ready for boosting
Example results
Summary
Database Classification using Support Vector Machines
Database classification
Data classification in statistics
Guidelines for classifying data
Common guidelines
Definitions
Definition and purpose of an SVM
The trick
Feature space and cheap computations
Drawing the line
More than classification
Downside
Reference resources
Predicting credit scores
Using R and an SVM to classify data in a database
Moving on
Summary
Database Structures and Machine Learning
Data structures and data models
Data structures
Data models
What's the difference?
Relationships
Machine learning
Overview of machine learning concepts
Key elements of machine learning
Representation
Evaluation
Optimization
Types of machine learning
Supervised learning
Unsupervised learning
Semi-supervised learning
Reinforcement learning
Most popular
Applications of machine learning
Machine learning in practice
Understanding
Preparation
Learning
Interpretation
Deployment
Iteration
Using R to apply machine learning techniques to a database
Understanding the data
Preparing
Data developer
Understanding the challenge
Cross-tabbing and plotting
Summary
Preface
Statistics are an absolute must prerequisite for any task in the area of data science but may also be the most feared deterrent for developers entering into the field of data science. This book will take you on a statistical journey from knowing very little to becoming comfortable using various statistical methods for typical data science tasks.
What this book covers
Chapter 1: Transitioning from Data Developer to Data Scientist, sets the stage for the transition from data developer to data scientist. You will understand the difference between a developer mindset versus a data scientist mindset, the important difference between the two, and how to transition into thinking like a data scientist.
Chapter 2: Declaring the Objectives, introduces and explains (from a developer’s perspective) the basic objectives behind statistics for data science and introduces you to the important terms and keys that are used in the field of data science.
Chapter 3: A Developer's Approach to Data Cleaning, discusses how a developer might understand and approach the topic of data cleaning using common statistical methods.
Chapter 4: Data Mining and the Database Developer, introduces the developer to mining data using R. You will understand what data mining is, why it is important, and feel comfortable using R for the most common statistical data mining methods: dimensional reduction, frequent patterns, and sequences.
Chapter 5: Statistical Analysis for the Database Developer, discusses the difference between data analysis or summarization and statistical data analysis and will follow the steps for successful statistical analysis of data, describe the nature of data, explore the relationships presented in data, create a summarization model from data, prove the validity of a model, and employ predictive analytics on a developed model.
Chapter 6: Database Progression to Database Regression, sets out to define statistical regression concepts and outline how a developer might use regression for simple forecasting and prediction within a typical data development project.
Chapter 7: Regularization for Database Improvement, introduces the developer to the idea of statistical regularization to improve data models. You will review what statistical regularization is, why it is important, and various statistical regularization methods.
Chapter 8: Data Development and Assessment, covers the idea of data model assessment and using statistics for assessment. You will understand what statistical assessment is, why it is important, and use R for statistical assessment.
Chapter 9: Databases and Neural Networks, defines the neural network model and draws from a developer’s knowledge of data models to help understand the purpose and use of neural networks in data science.
Chapter 10: Boosting and your Database, introduces the idea of using statistical boosting to better understand data in a database.
Chapter 11: Database Classification using Support Vector Machines, uses developer terminologies to define an SVM, identify various applications for its use and walks through an example of using a simple SVM to classify data in a database
Chapter 12: Database Structures and Machine Learning, aims to provide an explanation of the types of machine learning and shows the developer how to use machine learning processes to understand database mappings and identify patterns within the data.
What you need for this book
This book is intended for those with a data development background who are interested in possibly entering the field of data science and are looking for concise information on the topic of statistics with the help of insightful programs and simple explanation. Just bring your data development experience and an open mind!
Who this book is for
This book is intended for those developers who are interested in entering the field of data science and are looking for concise information on the topic of statistics with the help of insightful programs and simple explanation.
Conventions
In this book, you will find a number of text styles that distinguish between different kinds of information. Here are some examples of these styles and an explanation of their meaning.
Code words in text, database table names, folder names, filenames, file extensions, pathnames, dummy URLs, user input, and Twitter handles are shown as follows: In statistics, a boxplot is a simple way to gain information regarding the shape, variability, and center (or median) of a statistical data set, so we'll use the boxplot with our data to see if we can identify both the median Coin-in and if there are any outliers.
A block of code is set as follows:
MyFile <-"C:/GammingData/SlotsResults.csv" MyData <- read.csv(file=MyFile, header=TRUE, sep=",")
New terms and important words are shown in bold.
Reader feedback
Feedback from our readers is always welcome. Let us know what you think about this book-what you liked or disliked. Reader feedback is important for us as it helps us develop titles that you will really get the most out of. To send us general feedback, simply email [email protected], and mention the book's title in the subject of your message. If there is a topic that you have expertise in and you are interested in either writing or contributing to a book, see our author guide at www.packtpub.com/authors.
Customer support
Now that you are the proud owner of a Packt book, we have a number of things to help you to get the most from your purchase.
Downloading the example code
You can download the example code files for this book from your account at http://www.packtpub.com. If you purchased this book elsewhere, you can visit http://www.packtpub.com/support and register to have the files emailed directly to you. You can download the code files by following these steps:
Log in or register to our website using your email address and password.
Hover the mouse pointer on the
SUPPORT
tab at the top.
Click on
Code Downloads & Errata
.
Enter the name of the book in the
Search
box.
Select the book for which you're looking to download the code files.
Choose from the drop-down menu where you purchased this book from.
Click on
Code Download
.
Once the file is downloaded, please make sure that you unzip or extract the folder using the latest version of:
WinRAR / 7-Zip for Windows
Zipeg / iZip / UnRarX for Mac
7-Zip / PeaZip for Linux
The code bundle for the book is also hosted on GitHub at https://github.com/PacktPublishing/Statistics-for-Data-Science. We also have other code bundles from our rich catalogue of books and videos available at https://github.com/PacktPublishing/. Check them out!
Downloading the color images of this book
We also provide you with a PDF file that has color images of the screenshots/diagrams used in this book. The color images will help you better understand the changes in the output. You can download this file from https://www.packtpub.com/sites/default/files/downloads/StatisticsforDataScience_ColorImages.pdf.
Errata
Although we have taken every care to ensure the accuracy of our content, mistakes do happen. If you find a mistake in one of our books-maybe a mistake in the text or the code-we would be grateful if you could report this to us. By doing so, you can save other readers from frustration and help us improve subsequent versions of this book. If you find any errata, please report them by visiting http://www.packtpub.com/submit-errata, selecting your book, clicking on the Errata Submission Form link, and entering the details of your errata. Once your errata are verified, your submission will be accepted and the errata will be uploaded to our website or added to any list of existing errata under the Errata section of that title. To view the previously submitted errata, go to https://www.packtpub.com/books/content/support and enter the name of the book in the search field. The required information will appear under the Errata section.
Piracy
Piracy of copyrighted material on the internet is an ongoing problem across all media. At Packt, we take the protection of our copyright and licenses very seriously. If you come across any illegal copies of our works in any form on the internet, please provide us with the location address or website name immediately so that we can pursue a remedy. Please contact us at [email protected] with a link to the suspected pirated material. We appreciate your help in protecting our authors and our ability to bring you valuable content.
Questions
If you have a problem with any aspect of this book, you can contact us at [email protected], and we will do our best to address the problem.
Transitioning from Data Developer to Data Scientist
In this chapter (and throughout all of the chapters of this book), we will chart your course for starting and continuing the journey from thinking like a data developer to thinking like a data scientist.
Using developer terminologies and analogies, we will discuss a developer's objectives, what a typical developer mindset might be like, how it differs from a data scientist's mindset, why there are important differences (as well as similarities) between the two and suggest how to transition yourself into thinking like a data scientist. Finally, we will suggest certain advantages of understanding statistics and data science, taking a data perspective, as well as simply thinking like a data scientist.
In this chapter, we've broken things into the following topics:
The objectives of the data developer role
How a data developer thinks
The differences between a data developer and a data scientist
Advantages of thinking like a data scientist
The steps for transitioning into a data scientist mindset
So, let's get started!
Data developer thinking
Having spent plenty of years wearing the hat of a data developer, it makes sense to start out here with a few quick comments about data developers.
In some circles, a database developer is the equivalent of a data developer. But whether data or database, both would usually be labeled as an information technology (IT) professional. Both spend their time working on or with data and database technologies.
Your typical data developer will primarily be involved with creating and maintaining access to data rather than consuming that data. He or she will have input in or may make decisions on, choosing programming languages for accessing or manipulating data. We will make sure that new data projects adhere to rules on how databases store and handle data, and we will create interfaces between data sources.
In addition, some data developers are involved with reviewing and tuning queries written by others and, therefore, must be proficient in the latest tuning techniques, various query languages such as Structured Query Language (SQL), as well as how the data being accessed is stored and structured.
In summary, at least strictly from a data developer's perspective, the focus is all about access to valuable data resources rather than the consumption of those valuable data resources.
Objectives of a data developer
Every role, position, or job post will have its own list of objectives, responsibilities, or initiatives.
As such, in the role of a data developer, one may be charged with some of the following responsibilities:
Maintaining the integrity of a database and infrastructure
Monitoring and optimizing to maintain levels of responsiveness
Ensuring quality and integrity of data resources
Providing appropriate levels of support to communities of users
Enforcing security policies on data resources
As a data scientist, you will note somewhat different objectives. This role will typically include some of the objectives listed here:
Mining data from disparate sources
Identifying patterns or trending
Creating statistical models—modeling
Learning and assessing
Identifying insights and predicting
Do you perhaps notice a theme beginning here?
Note the keywords:
Maintaining
Monitoring
Ensuring
Providing
Enforcing
These terms imply different notions than those terms that may be more associated with the role of a data scientist, such as the following:
Mining
Trending
Modeling
Learning
Predicting
There are also, of course, some activities performed that may seem analogous to both a data developer and a data scientist and will be examined here.
Querying or mining
As a data developer, you will almost always be in the habit of querying data. Indeed, a data scientist will query data as well. So, what is data mining? Well, when one queries data, one expects to ask a specific question. For example, you might ask, What was the total number of daffodils sold in April? expecting to receive back a known, relevant answer such as in April, daffodil sales totaled 269 plants.
With data mining, one is usually more absorbed in the data relationships (or the potential relationships between points of data, sometimes referred to as variables) and cognitive analysis. A simple example might be: how does the average daily temperature during the month affect the total number of daffodils sold in April?
Another important distinction between data querying and data mining is that queries are typically historic in nature in that they are used to report past results (total sales in April), while data mining techniques can be forward thinking in that through the use of appropriate statistical methods, they can infer a future result or provide the probability that a result or event will occur. For example, using our earlier example, we might predict higher daffodil sales when the average temperature rises within the selling area.
Data quality or data cleansing
Do you think a data developer is interested in the quality of data in a database? Of course, a data developer needs to care about the level of quality of the data they support or provide access to. For a data developer, the process of data quality assurance (DQA) within an organization is more mechanical in nature, such as ensuring data is current and complete and stored in the correct format.
With data cleansing, you see the data scientist put more emphasis on the concept of statistical data quality. This includes using relationships found within the data to improve the levels of data quality. As an example, an individual whose age is nine, should not be labeled or shown as part of a group of legal drivers in the United States incorrectly labeled data.
Data modeling
Data developers create designs (or models) for data by working closely with key stakeholders based on given requirements such as the ability to rapidly enter sales transactions into an organization's online order entry system. During model design, there are three kinds of data models the data developer must be familiar with—conceptual, logical, and physical—each relatively independent of each other.
Data scientists create models with the intention of training with data samples or populations to identify previously unknown insights or validate current assumptions.
Issue or insights
A lot of a data developer's time may be spent monitoring data, users, and environments, looking for any indications of emerging issues such as unexpected levels of usage that may cause performance bottlenecks or outages. Other common duties include auditing, application integrations, disaster planning and recovery, capacity planning, change management, database software version updating, load balancing, and so on.
Data scientists spend their time evaluating and analyzing data, and information in an effort to discover valuable new insights. Hopefully, once established, insights can then be used to make better business decisions.
Thought process
Someone's mental procedures or cognitive activity based on interpretations, past experiences, reasoning, problem-solving, imagining, and decision making make up their way of thinking or their thought process.
One can only guess how particular individuals will actually think, or their exact thoughts at a given point of time or during an activity, or what thought process they will use to accomplish their objectives, but in general terms, a data developer may spend more time thinking about data convenience (making the data available as per the requirements), while data scientists are all about data consumption (concluding new ways to leverage the data to find insights into existing issues or new opportunities).
To paint a clearer picture, you might use the analogy of the auto mechanic and the school counselor.
An auto mechanic will use his skills along with appropriate tools to keep an automobile available to its owner and running well, or if there has been an issue identified with a vehicle, the mechanic will perform diagnosis for the symptoms presented and rectify the problem. This is much like the activities of a data developer.
With a counselor, he or she might examine a vast amount of information regarding a student's past performance, personality traits, as well as economic statistics to determine what opportunities may exist in a particular student's future. In addition, multiple scenarios may be studied to predict what the best outcomes might be, based on this individual student's resources.
Clearly, both aforementioned individuals provide valuable services but use (maybe very) different approaches and individual thought processes to produce the desired results.
Although there is some overlapping, when you are a data developer, your thoughts are normally around maintaining convenient access to appropriate data resources but not particularly around the data's substance, that is, you may care about data types, data volumes, and accessibility paths but not about whether or what cognitive relationships exist or the powerful potential uses for the data.
In the next section, we will explore some simple circumstances in an effort to show various contrasts between the data developer and the data scientist.
Developer versus scientist
To better understand the differences between a data developer and data scientist, let's take a little time here and consider just a few hypotheticals (yet still realistic) situations that may occur during your day.
New data, new source
What happens when new data or a new data source becomes available or is presented?
Here, new data usually means that more current or more up-to-date data has become available. An example of this might be receiving a file each morning of the latest month-to-date sales transactions, usually referred to as an actual update.
You may receive files of data periodically from an online transactions processing system, which provide the daily sales or sales figures from the first of the month to the current date. You'd want your business reports to show the total sales numbers that include the most recent sales transactions.
The idea of a new data source is different. If we use the same sort of analogy as we used previously, an example of this might be a file of sales transactions from a company that a parent company newly acquired. Perhaps another example would be receiving data reporting the results of a recent online survey. This is the information that's collected with a specific purpose in mind and typically is not (but could be) a routine event.
In the previous scenarios, a data developer would most likely be (should be) expecting updated files and have implemented the Extract, Transform, and Load (ETL) processes to automatically process the data, handle any exceptions, and ensure that all the appropriate reports reflect the latest, correct information. Data developers would also deal with transitioning a sales file from a newly acquired company but probably would not be a primary resource for dealing with survey results (or the 1000 Genomes Project).
Data scientists are not involved in the daily processing of data (such as sales) but will be directly responsible for a survey results project. That is, the data scientist is almost always hands-on with initiatives such as researching and acquiring new sources of information for projects involving surveying. Data scientists most likely would have input even in the designing of surveys as they are the ones who will be using that data in their analysis.
Quality questions
Suppose there are concerns about the quality of the data to be, or being, consumed by the organization. As we eluded to earlier in this chapter, there are different types of data quality concerns such as what we called mechanical issues as well as statistical issues (and there are others).
If management is questioning the validity of the total sales listed on a daily report or perhaps doesn't trust it because the majority of your customers are not legally able to drive in the United States, the number of the organizations repeat customers are declining, you have a quality issue:
Quality is a concern to both the data developer and the data scientist. A data developer focuses more on timing and formatting (the mechanics of the data), while the data scientist is more interested in the data's statistical quality (with priority given to issues with the data that may potentially impact the reliability of a particular study).
Querying and mining
Historically, the information technology group or department has been beseeched by a variety of business users to produce and provide reports showing information stored in databases and systems that are of interest.
These ad hoc reporting requests have evolved into requests for on-demand raw data extracts (rather than formatted or pretty printed reports) so that business users could then import the extracted data into a tool such as MS Excel (or others), where they could then perform their own formatting and reporting, or perform further analysis and modeling. In today's world, business users demand more self-service (even mobile) abilities to meet their organization's (or an individual's) analytical and reporting needs, expecting to have access to the updated raw data stores, directly or through smaller, focus-oriented data pools.
Creating ad hoc reports and performing extracts based on specific on-demand needs or providing self-service access to data falls solely to the role of the organization's data developer. However, take note that a data scientist will want to periodically perform his or her own querying and extracting—usually as part of a project they are working on. They may use these query results to determine the viability and availability of the data they need or as part of the process to create a sampling or population for specific statistical projects. This form of querying may be considered to be a form of data mining and goes much deeper into the data than queries might. This work effort is typically performed by a data scientist rather than a data developer.
Performance
You can bet that pretty much everyone is, or will be, concerned with the topic of performance. Some forms (of performance) are perhaps a bit more quantifiable, such as what is an acceptable response time for an ad hoc query or extract to complete? Or perhaps what are the total number of mouse-clicks or keystrokes required to enter a sales order? Others may be a bit more difficult to answer or address, such as why does it appear that there is a downward trend in the number of repeat customers?
It is the responsibility of the data developer to create and support data designs (even be involved with infrastructure configuration options) that consistently produce swift response times and are easy to understand and use.
