Sunday, February 28, 2021
  • Setup menu at Appearance » Menus and assign menu to Top Bar Navigation
Advertisement
  • AI Development
    • Artificial Intelligence
    • Machine Learning
    • Neural Networks
    • Learn to Code
  • Data
    • Blockchain
    • Big Data
    • Data Science
  • IT Security
    • Internet Privacy
    • Internet Security
  • Marketing
    • Digital Marketing
    • Marketing Technology
  • Technology Companies
  • Crypto News
No Result
View All Result
NikolaNews
  • AI Development
    • Artificial Intelligence
    • Machine Learning
    • Neural Networks
    • Learn to Code
  • Data
    • Blockchain
    • Big Data
    • Data Science
  • IT Security
    • Internet Privacy
    • Internet Security
  • Marketing
    • Digital Marketing
    • Marketing Technology
  • Technology Companies
  • Crypto News
No Result
View All Result
NikolaNews
No Result
View All Result
Home Data Science

Calculating Price Elasticity Through KNIME

September 15, 2019
in Data Science
Calculating Price Elasticity Through KNIME
587
SHARES
3.3k
VIEWS
Share on FacebookShare on Twitter

As per Wikipedia, Price Elasticity of Demand (PED or ED or PE) is a measure used in economics to show the responsiveness, or change, of the quantity demanded of a good or service to a change in its price when nothing but the price changes. In more precise business terms, it helps in finding those products which have their sales more/less susceptible to price changes. As we know, the demand is inversely proportional to price, it is quite imperative to know this information for optimising sales and margins.

Without going much into literature of Price Elasticity (PE), I shall explain how I implemented ‘no-code-analytics’ through KNIME to get the PE values from real sales and price data. I also calculated Cross Price Elasticity (CPE) since I had competitor prices at my disposal too. I shall be highlighting all the modules that I used along with the respective input-output.

You might also like

The Education Industrial Complex: The Hammer We Have

Increasing Adoption of Informatics will Promote Growth of Data Analytics Outsourcing Market

The Ethereum Virtual Machine (EVM)

As in any analytics project, here also the major weight lifting is data engineering – getting the data in shape for the final model to run. I started with the two Excel Reader (1) modules to ingest price and sales data. The first thing I did was drop couple of unnecessary columns like county, company, model name (I will use part numbers to identify a product), short description, list price, discount etc. using the module Column Filter (2). The very first column I engineered was date column, to remove time stamp and filter just date using the Date & Time to String (3) module. Now I removed unnecessary rows with junk values using the Row Filter (4.1) module. If a dataset requires special treatment, you can filter those rows and then append to the main data using Concatenate (4.2) module post processing. Now comes the interesting part, converting the long format data to wide. I need only one row for all the prices (across competing retailers) a product had on a particular day. Hence, as an output, I will have all the retailers in the data in respective columns, with price on that particular day as value. For this I used Pivoting (5) module with part number & date as grouping columns, retailer will go into columns and price would be the value selected. It doesn’t allow you to aggregate multiple values and hence you have to use the Column Aggregator (6) module. There is flexibility of selecting the aggregation mode. In my case, I selected the minimum value so that I can capture the most competitive seller price on a marketplace website. Now, it is time to join price data with sales data (which was being processed in parallel). The join was achieved using Joiner (7) module. You can select the join mode (inner, outer etc.), which columns to be used to join and which columns are to be kept in the final output. If you want to do any kind of manipulation on date e.g. extracting day/year, changing the format, getting the week number etc., you can do it by modules Extract Date & Time Fields (8.1), Number To String (8.2) and String Manipulation (8.3).

Now we try to implement a regression model on every sales X price combination for a part number across the time period. So, we have to loop on every part number and run a regression model for each part number. I must say, doing this is much easier in an R or Python compared to KNIME.  Start with a Group Loop Start (9.1) module. Then comes the Linear regression Learner (9.2) to run the model on single part number. But since it doesn’t provide an R2 value, you would have to take a circuitous route involving Regression Predictor (9.3) with outputs from Loop (data) & Regression learner (equation) as input. Numeric Scorer (9.4) will provide the R2 value, Constant Value Column (9.5) shall add the loop iteration identifier column to data to later join with the coefficients coming from Regression Learner (9.2). Post joining the Coefficients and R2 values using Joiner (9.8), loop can be closed using Loop End (9.9). In between, there are also two important stages – RowID (9.6): helps in identifying the R2 value among multiple values and then Row Filter (9.7): helps in keeping only R2 value and removing rest of statistics.  

Use a Joiner (10) to combine sales, prices, coefficients and R2 values in one table. Now, we can now use the module Math Formula Multi Column (11) to apply the formulas:

Sale = Intercept + A*Own Price + B*Competitor Price  

Own Price Elasticity  =  A* (average Own price/average Own sales)

Cross Price Elasticity = B*(average Competitor Price/average Own sales)

Fig 1: The arrangement of modules to run regression in loop and get coefficients/R2 values

Fig 2: The output of Numeric Scorer

Fig 3: The output from Constant Value Column (9.5)

Fig 4: The output from RowID (9.6)

Fig 5: Complete Flow


Credit: Data Science Central By: saurabh ajmera

Previous Post

Modern Warfare Beta Boldly Abandons Minimap, But Not for Long

Next Post

Most consumers will refuse to work with enterprises that won’t keep their data secure

Related Posts

The Education Industrial Complex: The Hammer We Have
Data Science

The Education Industrial Complex: The Hammer We Have

February 27, 2021
Increasing Adoption of Informatics will Promote Growth of Data Analytics Outsourcing Market
Data Science

Increasing Adoption of Informatics will Promote Growth of Data Analytics Outsourcing Market

February 27, 2021
The Ethereum Virtual Machine (EVM)
Data Science

The Ethereum Virtual Machine (EVM)

February 27, 2021
Levels of Measurement (Nominal, Ordinal, Interval, Ratio) in Statistics
Data Science

Levels of Measurement (Nominal, Ordinal, Interval, Ratio) in Statistics

February 27, 2021
Give Your Business Users Simple Augmented Analytics
Data Science

Give Your Business Users Simple Augmented Analytics

February 26, 2021
Next Post
Most consumers will refuse to work with enterprises that won’t keep their data secure

Most consumers will refuse to work with enterprises that won’t keep their data secure

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Recommended

Plasticity in Deep Learning: Dynamic Adaptations for AI Self-Driving Cars

Plasticity in Deep Learning: Dynamic Adaptations for AI Self-Driving Cars

January 6, 2019
Microsoft, Google Use Artificial Intelligence to Fight Hackers

Microsoft, Google Use Artificial Intelligence to Fight Hackers

January 6, 2019

Categories

  • Artificial Intelligence
  • Big Data
  • Blockchain
  • Crypto News
  • Data Science
  • Digital Marketing
  • Internet Privacy
  • Internet Security
  • Learn to Code
  • Machine Learning
  • Marketing Technology
  • Neural Networks
  • Technology Companies

Don't miss it

TikTok agrees to pay $92 million to settle teen privacy class-action lawsuit
Internet Security

TikTok agrees to pay $92 million to settle teen privacy class-action lawsuit

February 28, 2021
Machine Learning as a Service (MLaaS) Market 2020 Emerging Trend and Advancement Outlook 2025
Machine Learning

Key Company Profile, Production Revenue, Product Picture and Specifications 2025

February 28, 2021
Cybercrime groups are selling their hacking skills. Some countries are buying
Internet Security

Cybercrime groups are selling their hacking skills. Some countries are buying

February 28, 2021
New AI Machine Learning Reduces Mental Health Misdiagnosis
Machine Learning

Machine Learning May Reduce Mental Health Misdiagnosis

February 28, 2021
Why would you ever trust Amazon’s Alexa after this?
Internet Security

Why would you ever trust Amazon’s Alexa after this?

February 28, 2021
AI & ML Are Not Same. Here's Why – Analytics India Magazine
Machine Learning

AI & ML Are Not Same. Here's Why – Analytics India Magazine

February 27, 2021
NikolaNews

NikolaNews.com is an online News Portal which aims to share news about blockchain, AI, Big Data, and Data Privacy and more!

What’s New Here?

  • TikTok agrees to pay $92 million to settle teen privacy class-action lawsuit February 28, 2021
  • Key Company Profile, Production Revenue, Product Picture and Specifications 2025 February 28, 2021
  • Cybercrime groups are selling their hacking skills. Some countries are buying February 28, 2021
  • Machine Learning May Reduce Mental Health Misdiagnosis February 28, 2021

Subscribe to get more!

© 2019 NikolaNews.com - Global Tech Updates

No Result
View All Result
  • AI Development
    • Artificial Intelligence
    • Machine Learning
    • Neural Networks
    • Learn to Code
  • Data
    • Blockchain
    • Big Data
    • Data Science
  • IT Security
    • Internet Privacy
    • Internet Security
  • Marketing
    • Digital Marketing
    • Marketing Technology
  • Technology Companies
  • Crypto News

© 2019 NikolaNews.com - Global Tech Updates