ETL Pipeline Documentation: Here are my Tips and Tricks!

ETL pipeline documentation is necessary on all team data projects. In fact, I make ETL pipeline documentation every time I have a project where I am doing ETL, even if I’m the only one doing it. Extract, transform, and load (ETL) protocols are typically quite complex. Even what initially looks like a simple process of transforming three native variables into nine analytic variables – as I demonstrate in this video – turns out to be actually rather complex.

Watch my video to get a demonstration on how to document your ETL pipeline

ETL Pipeline Documentation: Different Types

In the video, I classify ETL pipeline documentation into two categories: the minimum necessary, and additional helpful files that can be added. What I list under the “minimum necessary” is basically two items:

A complete data dictionary, with well-documented picklists and available crosswalks, and
A copy of the final analytic dataset.

You can download all of these files from GitHub.

Data Dictionary

The main part of a data dictionary provides a list of names of variables in a data table, and provides necessary information about them, such as what they mean, how they are populated, and in what format they are stored. Picklists are like dropdown lists, where a value refers to a level in a categorical variable (e.g., 1 = Yes, 2 = No, 9 = Don’t Know). Crosswalks are like picklists on steroids – meaning they are very extensive picklists that can be joined to the main dataset to add other columns to enhance its analytic utility.

Analytic Datasets for Research

When you engage in ETL, you are transforming the native extract you received from the data provider into an analytic dataset on which you can base your analysis for research. Through discrete transformation steps, your ETL code transforms the native extract into the analytic dataset.

Other ETL Pipeline Documentation

Other documentation that can be helpful is:

Well-organized ETL code
An ETL pipeline diagram
Example deidentified source data.

Well-organized ETL Code

Well-organized ETL code is modular, and includes only one or a few transformation steps per code file. The code files use naming conventions that make the code line up in order of operation. Other naming conventions (e.g., for datasets and variables) are applied. To learn more about this, read Chapter 4 of my book, “Mastering SAS Programming for Data Warehousing”.

ETL Pipeline Diagram

This ETL pipeline diagram is basically a map that summarizes the code. It goes hand in hand with the code and the transformation steps. If you have trouble with a particular variable in the analytic dataset, this diagram can help you troubleshoot which code to open and modify. Well-organized code and diagrams like these are necessary if you want to automate your ETL code using SAS or other programs. I go over ETL code automation in Chapter 8 of my SAS book. The videos below from SAS provide an example of doing this in SAS Data Integration Studio.

SAS provided this video to show you how to automate your ETL pipeline in SAS DI Studio.

In SAS Data Integration Studio, a job refers to a pipeline you can create, save, and come back to edit.

Example of Deidentified Source Data

Sometimes if you get confused about data, you can ask your data provider to make up a few fake records that follow validation rules for you.

These data are obviously faked – but they serve to show the receiver of this faked data the format, structure, and some business rules of the data. Receiving a faked dataset like this – even if it is only a few rows – can give you the opportunity to start building transformation code (especially complex date handling), even before you receive the real data from the provider.

Updated February 22, 2023. Revised Banners June 18, 2023.

Read all of our data science blog posts!

You can do a population-based analysis if the original dataset used multi-stage sampling.

Apply Weights? It’s Easy in R with the Survey Package!

Apply weights to get weighted proportions and counts! Read my blog post to learn how [...]

11
Nov

You can make mean, median and mode with a continuous variable.

Make Categorical Variable Out of Continuous Variable

Make categorical variables by cutting up continuous ones. But where to put the boundaries? Get [...]

04
Nov

Remove Rows in R with the Subset Command

Remove rows by criteria is a common ETL operation – and my blog post shows [...]

28
Oct

The open government data movements in many countries have resulted in government data being available online.

CDC Wonder for Studying Vaccine Adverse Events: The Shameful State of US Open Government Data

CDC Wonder is an online query portal that serves as a gateway to many government [...]

24
Jun

If you are a data scientist, you may want to do statistics, but you may also be interested in machine learning and artificial intelligence.

AI Careers: Riding the Bubble

AI careers are not easy to navigate. Read my blog post for foolproof advice for [...]

16
Jun

The Black Friday Death Count database has a list of news reports of deaths or injuries on Black Friday.

Descriptive Analysis of Black Friday Death Count Database: Creative Classification

Descriptive analysis of Black Friday Death Count Database provides an example of how creative classification [...]

27
Nov

What if you have too many categories in a categorical variable? Your cardinality is too high for a chi-square analysis.

Classification Crosswalks: Strategies in Data Transformation

Classification crosswalks are easy to make, and can help you reduce cardinality in categorical variables, [...]

24
Nov

Want to learn more about pharmacy data? You can use adverse event data in a data science portfolio project.

FAERS Data: Getting Creative with an Adverse Event Surveillance Dashboard

FAERS data are like any post-market surveillance pharmacy data – notoriously messy. But if you [...]

22
Nov

If you work on a big data project with multiple source datasets, you run the risk of forgetting exactly how you blended them together.

Dataset Source Documentation: Necessary for Data Science Projects with Multiple Data Sources

Dataset source documentation is good to keep when you are doing an analysis with data [...]

09
Nov

In base R, you can execute SQL-like joins, as long as you use the correct code syntax.

Joins in Base R: Alternative to SQL-like dplyr

Joins in base R must be executed properly or you will lose data. Read my [...]

07
Nov

If you are interested in population-level surveillance data, you might have thought about using NHANES data in portfolio projects.

NHANES Data: Pitfalls, Pranks, Possibilities, and Practical Advice

NHANES data piqued your interest? It’s not all sunshine and roses. Read my blog post [...]

02
Nov

When using big data, you will want to make visualizations. How do you use color to the greatest communicative advantage?

Color in Visualizations: Using it to its Full Communicative Advantage

Color in visualizations of data curation and other data science documentation can be used to [...]

09
Oct

The defaults in PowerPoint are really set up for making presentations, not data visualizations.

Defaults in PowerPoint: Setting Them Up for Data Visualizations

Defaults in PowerPoint are set up for slides – not data visualizations. Read my blog [...]

06
Oct

Adding text and arrows to diagrams can help your audience navigate the image, and understand what you are trying to communicate.

Text and Arrows in Dataviz Can Greatly Improve Understanding

Text and arrows in dataviz, if used wisely, can help your audience understand something very [...]

04
Oct

If you use good judgment in choosing chapes and images to add to your data visualizations, your audience will be enlightened.

Shapes and Images in Dataviz: Making Choices for Optimal Communication

Shapes and images in dataviz, if chosen wisely, can greatly enhance the communicative value of [...]

01
Oct

When you use a data analysis program like R or SAS, you often have to do some data editing. It can be difficult because the software was intended for calculations, not transformation.

Table Editing in R is Easy! Here Are a Few Tricks…

Table editing in R is easier than in SAS, because you can refer to columns, [...]

24
Aug

Logistic regression calculate the log odds of the probability of the outcome. Many people are used to using SAS for logistic regression, but you can also use R.

R for Logistic Regression: Example from Epidemiology and Biostatistics

R for logistic regression in health data analytics is a reasonable choice, if you know [...]

1 Comments

17
Aug

Did you know it is possible to integrate SAS with other data environments, like Microsoft SQL or Excel?

Connecting SAS to Other Applications: Different Strategies

Connecting SAS to other applications is often necessary, and there are many ways to do [...]

30
Jul

Are you a data scientist who is interested in doing independent portfolio projects to sharpen your skills? Then I strongly suggest you get a coach or a mentor.

Portfolio Project Examples for Independent Data Science Projects

Portfolio project examples are sometimes needed for newbies in data science who are looking to [...]

13
Jul

If you are a health data analyst or a biostatistician, we might find computer programmers and application developers use different terminology for the same ideas and concepts.

Project Management Terminology for Public Health Data Scientists

Project management terminology is often used around epidemiologists, biostatisticians, and health data scientists, and it’s [...]

25
Jun

If you work on front-ends or back-ends of health applications, you are probably already familiar with the concepts of Agile and rapid application development.

Rapid Application Development Public Health Style

“Rapid application development” (RAD) refers to an approach to designing and developing computer applications. In [...]

22
Jun

Data systems started being in use in the 1960s and 1970s, but these were flat systems, usually using IBM mainframes.

Understanding Legacy Data in a Relational World

Understanding legacy data is necessary if you want to analyze datasets that are extracted from [...]

19
Jun

How the front-end and back-end are connected can impact how data are stored in the application. So if you extract the data, you can have data quality problems caused by the front-end.

Front-end Decisions Impact Back-end Data (and Your Data Science Experience!)

Front-end decisions are made when applications are designed. They are even made when you design [...]

16
Jun

Slow queries can happen in SAS, R, Python, SQL or any database language. These slow queries have a cost.

Reducing Query Cost (and Making Better Use of Your Time)

Reducing query cost is especially important in SAS – but do you know how to [...]

13
Jun

If you need data to do a project, read this blog post for information.

Curated Datasets: Great for Data Science Portfolio Projects!

Curated datasets are useful to know about if you want to do a data science [...]

31
May

Public health, artificial intelligence, and data science trivia! Fun! Educational! Test your knowledge!

Statistics Trivia for Data Scientists

Statistics trivia for data scientists will refresh your memory from the courses you’ve taken – [...]

12
Apr

When working in data science, there are some tips and tricks to managing your communication and relationship with superiors that can help you advance in your career.

Management Tips for Data Scientists

Management tips for data scientists can be used by anyone – at work and in [...]

23
Mar

REDCap mess on your hands? The REDCap designers made the application so loosey goosey, you can really program yourself into a messy corner if you don't plan well.

REDCap Mess: How it Got There, and How to Clean it Up

REDCap mess happens often in research shops, and it’s an analysis showstopper! Read my blog [...]

07
Mar

If you are an aspiring data scientist, you will need to know how GitHub works. You will probably want to use it for your projects.

GitHub Beginners in Data Science: Here’s an Easy Way to Start!

GitHub beginners – even in data science – often feel intimidated when starting their GitHub [...]

28
Feb

This blog post shows you how to properly document your extract, transform, and load code.

ETL Pipeline Documentation: Here are my Tips and Tricks!

ETL pipeline documentation is great for team communication as well as data stewardship! Read my [...]

22
Feb

How do you measure how long it takes for code to run in different programs? And why would you want to measure something like that? Mainly, the reason to benchmark runtime is so that you can figure out how to optimize your code.

Benchmarking Runtime is Different in SAS Compared to Other Programs

Benchmarking runtime is different in SAS compared to other programs, where you have to request [...]

13
Dec

What is an end-to-end AI pipeline? And why are academics so bad at making one? These are different ideas we will examine in this blog post.

End-to-End AI Pipelines: Can Academics Be Taught How to Do Them?

End-to-end AI pipelines are being created routinely in industry, and one complaint is that academics [...]

25
Nov

There are different ways to refer to variables in R dataframes. You can use a field names, and you can also use field numbers.

Referring to Columns in R by Name Rather than Number has Pros and Cons

Referring to columns in R can be done using both number and field name syntax. [...]

17
Oct

The paste command is used to concatenate strings in R. You can use it different ways, which is what I demonstrate in my blog and videos.

The Paste Command in R is Great for Labels on Plots and Reports

The paste command in R is used to concatenate strings. You can leverage the paste [...]

12
Oct

You do not need to use the default R colors on your plot. You don't even need to limit yourself to named colors on cheat sheets.

Coloring Plots in R using Hexadecimal Codes Makes Them Fabulous!

Recoloring plots in R? Want to learn how to use an image to inspire R [...]

11
Oct

Error bars on plots can provide the audience an estimate of the amount of certainty you have with your estimates.

Adding Error Bars to ggplot2 Plots Can be Made Easy Through Dataframe Structure

Adding error bars to ggplot2 in R plots is easiest if you include the width [...]

11
Oct

AI on the edge refers to doing the AI processing and equations at the site of the object collecting the data.

AI on the Edge: What it is, and Data Storage Challenges it Poses

“AI on the edge” was a new term for me that I learned from Marc [...]

06
Jun

How do you make a pie chart in ggplot2 package in R? It's not that obvious

Pie Chart ggplot Style is Surprisingly Hard! Here’s How I Did it

Pie chart ggplot style is surprisingly hard to make, mainly because ggplot2 did not give [...]

20
Apr

Time series plots can be customized if you use package ggplot2 in R. You can place labels and configure axes.

Time Series Plots in R Using ggplot2 Are Ultimately Customizable

Time series plots in R are totally customizable using the ggplot2 package, and can come [...]

11
Apr

It is possible to use data curation to solve the problem of a confusion vector containing options.

Data Curation Solution to Confusing Options in R Package UpSetR

Data curation solution that I posted recently with my blog post showing how to do [...]

08
Apr

If you are having trouble setting options using R making plots, then you should read this blog post.

Making Upset Plots with R Package UpSetR Helps Visualize Patterns of Attributes

Making upset plots with R package UpSetR is an easy way to visualize patterns of [...]

4 Comments

03
Apr

There are two main ways to make box plots in R, and this blog post shows you how, and explains the differences.

Making Box Plots Different Ways is Easy in R!

Making box plots in R affords you many different approaches and features. My blog post [...]

31
Mar

If you want to use R for a project and the source CSV is very big, it can improve input/output efficiency to convert the file to an RDS.

Convert CSV to RDS When Using R for Easier Data Handling

Convert CSV to RDS is what you want to do if you are working with [...]

25
Mar

This case example shows a use case where we estimated sample size in GPower under different conditions.

GPower Case Example Shows How to Calculate and Document Sample Size

GPower case example shows a use-case where we needed to select an outcome measure for [...]

18
Feb

Many data scientists interested in health are looking to query the Global Burden of Disease database, also known as the GHDx

Querying the GHDx Database: Demonstration and Review of Application

Querying the GHDx database is challenging because of its difficult user interface, but mastering it [...]

16
Feb

You need to come up with names of variables in SAS and in R, but they need to be compatible with both languages if you are running a data warehouse.

Variable Names in SAS and R Have Different Restrictions and Rules

Variable names in SAS and R are subject to different “rules and regulations”, and these [...]

04
Feb

When doing data processing, especially extract-transform-load (ETL) into a data warehouse, you might need to refer to the variables in your code, and it's done differently in SAS vs. R.

Referring to Variables in Processing Data is Different in SAS Compared to R

Referring to variables in processing is different conceptually when thinking about SAS compared to R. [...]

28
Jan

If you are a data scientist working with large datasets, you need to learn the commands to count both columns and rows in the dataset, whether you are using SAS or R.

Counting Rows in SAS and R Use Totally Different Strategies

Counting rows in SAS and R is approached differently, because the two programs process data [...]

14
Jan

Why use particular data formats for different programming languages in statistics? Because the programs can then process the data faster and with more accuracy.

Native Formats in SAS and R for Data Are Different: Here’s How!

Native formats in SAS and R of data objects have different qualities – and there [...]

07
Jan

You can use SAS and R together in one project. I show you how to develop an analytic dataset in R and put it in SAS ODA for analysis.

SAS-R Integration Example: Transform in R, Analyze in SAS!

Looking for a SAS-R integration example that uses the best of both worlds? I show [...]

26
Dec

This is an example of a dumbbell plot from the ggalt package in R that you can also use in RStudio

Dumbbell Plot for Comparison of Rated Items: Which is Rated More Highly – Harvard or the U of MN?

Want to compare multiple rankings on two competing items – like hotels, restaurants, or colleges? [...]

2 Comments

20
Sep

This is the forrest plot resulting from analysis with open source statistical software R using package rmeta.

14
Oct

The Likert package in R can visualize categorical data.

Doing Surveys? Try my R Likert Plot Data Hack!

I love the Likert package in R, and use it often to visualize data. The [...]

2 Comments

11
Oct

Epidemiologic Curve of 2015 Middle East Respiratory Virus Outbreak Using R EpiCurve Package

I Used the R Package EpiCurve to Make an Epidemiologic Curve. Here’s How It Turned Out.

With all this talk about “flattening the curve” of the coronavirus, I thought I would [...]

14
Mar

Little table of X Y Pairs with a Regresion Diagram with Least Squares Line

Which Independent Variables Belong in a Regression Equation? We Don’t All Agree, But Here’s What I Do.

During my failed attempt to get a PhD from the University of South Florida, my [...]

24
Aug

ETL pipeline documentation is great for team communication as well as data stewardship! Read my blog post to learn my tips and tricks.

ETL Pipeline Documentation: Different Types

throw out your books

…and learn statistics online!

Data Dictionary

Analytic Datasets for Research

advance your career

in data science

Other ETL Pipeline Documentation

Well-organized ETL Code

ETL Pipeline Diagram

Be an etl pro

Run a data system

Example of Deidentified Source Data

Monika Wahi

Leave a Reply Cancel reply