Oil and Gas Workshop | Applied Data Science and ML

Last updated:
9:15am Jun 26, 2020

Data science is a very demanded career that requires a mix of skills, including:

  • Accessing your data
  • Data wrangling, transformation, manipulation
  • Data analytics (identify trends and patterns from your data, gather valuable insights for your business)
  • Connect the dots (create visualizations and dashboards that tell a business story)
  • Push the analytics to the next level (apply ML algorithms with a user-friendly interface, put the algorithm behind the button for business users)
  • Connect the dots between technical users and non-technical users with a business purpose

This workshop is designed for anyone who has an interest in data and data science and does not require any experience with TIBCO Spotfire or machine learning.

Workshop Data: Scroll down this page and download the Spotfire Workshop.zip file to have access to all the data needed.

What If I don't have Spotfire?

Workshop Exercise

Key Objectives


Hands-on (Follow along with these videos/pause as needed)

Create a production dashboard from zero

(~45 mins)

• Import data 

• Link all sheets based on UWI

• Visualize key KPIs, production trends and geographical location of the Wells

• Calculate monthly operating rates in Mscf/d for each product (production volume/production hours) *24

• Record the UWI, the cumulative gas production, and the final operating gas rate for each Well


1.1 Import Data, KPI chart, Combination Chart (13:33)

1.2 Cross Table (7:15)

1.3 Add the Well length to the map (14:02)

1.4 Add a TMS* layer and a Text Area (5:58)

Note: The TMS Example URL is included as a .txt file on the Spotfire_workshop.zip available at the end of this page as an attachment.

1.5 Tips and tricks for good looking dashboards


GeoAnalytics example: Map Contour

(~15 mins)

In the Energy domain, different measurements might be gathered from a set of locations. The locations can be plotted on a map chart and in addition to visualizations such a bar graphs, heatmaps, etc. Contour lines on maps can be an excellent way of gaining visual insight into changes in measurements with respect to geography. We will cover:

• How to consume R/Python scripts into your Spotfire app

• How to enhance a map visualization with multiple layers


2.1 Map Contour (9:35)

Note: Starting URL here

Applied Machine learning: K-means and regression Modeling


Machine Learning algorithms such as Classification, Similarity, Clustering, Regressions, etc. allow the user to get a better more detailed and accurate grasp on the patterns displayed by the data. Spotfire users can use out-of-the-box statistical capabilities from the tools menu. These methods when combined with visualizations and maps allow the user to benefit from statistical data analysis without needing specialized domain knowledge of the same.


3.1 ML Unsupervised Example K-means clustering (5:09)

3.2 Regression Modeling (9:54)


Applied Machine Learning: DCA Example

(~20 mins)

Decline Curve Analysis (DCA) is a graphical procedure used for analyzing declining production rates and forecasting the future performance of oil and gas wells. Fitting a line through the performance history and assuming this same trend will continue in the future forms the basis of the DCA concept.

This example will show you how to consume a script (multiple programming languages are supported in Spotfire including R/Python) and use Spotfire as a visual interface to communicate the ML results.

• Put an algorithm behind the button.


4.1 DCA Example in Spotfire (13:56)

Note: Starting URL here



Keep sharping your data science skills and scale data science across your organization to solve complex challenges faster and speed innovation with TIBCO® Data Science, a comprehensive platform for operationalizing data science


• Complement your Spotfire training with this list of free training videos by topic

• Push Data Science to the next level with AutoML and pre-build ML algorithms

• Join Dr. Spotfire webinars and ask questions live. List of previous topics here

• Join Dr. Data Science webinars and ask questions live


Note: All the data attached as part of the .zip folder was created with the purpose of this exercise.


Package icon spotfire_workshop.zip1.05 MB