Hands-On Advanced Flux Training

November 8-9, 2022 / 9 am – 5 pm GMT

Duration: Two-day, hands-on in-person course.
Date: November 8-9, 2022 | 9 am – 5 pm GMT
Location: Avonmouth House, 6 Avonmouth Street, London, SE1 6NX, UK
Available spots: Spots are limited and only the first 50 will be able to register. All students who successfully complete the course will earn a sharable digital badge in recognition of their accomplishment.
Registration: The fee for Flux Training is £500 (Registration for the training will also include automatic registration to the InfluxDays virtual event on November 2-3, 2022). If you need assistance with your registration, please contact [email protected].

“The last Flux training I attended by this team was really useful. The examples and exercises they provided were really helpful in deepening our understanding of the data structures at play, and much aligned to the spirit of the InfluxDays conference itself where the attendees were encouraged to demonstrate & help each other further their learning.”

Jamie Milton, Senior Analytics Consultant, Fort Digital

Training overview

This instructor-led live in-person course from InfluxDB University dives into advanced uses of the Flux language for InfluxDB 2.x by InfluxData. The course will cover subjects from Data Engineering to Data Science using Flux to manipulate time series data.

After taking this class, attendants will be able to:

  • Articulate and implement advanced use cases for InfluxDB
  • Explore time series schemata understanding the measurements, the tag sets and the field sets
  • Understand alternatives in modeling time series schemata
  • Build data enrichment pipelines that compute metrics across multiple time series using Fluxlang
  • Understand how to choose between joining and pivoting time series
  • Explore time series data to detect trends using smoothing operations and learn to detrend a time series
  • Dissect time series data to identify how data spread evolves over time using aggregate windows and learn to normalize a time series
  • Look for correlation over time between pairs of time series and assess it numerically
  • Recognize repetitive patterns in a time series (a.k.a., seasonality) and learn to remove it
  • Understand the basics of time series forecasting
  • Learn to assess if it is possible to forecast a time series
  • Practically use the Holt-Winters method to forecast a time series that presents both trend and seasonality
  • Visualize the query results using advanced graphs from InfluxDB Cloud UI
  • Learn to develop dashboards that users can customize using Variables
  • Build and deploy real-time data analytics pipelines

Audience

Data engineers who want to deepen their understanding of Fluxlang to enhance their ability to manipulate time series data and provide real-time analytics.

Prerequisites and notes

  • Basic notions of InfluxDB data model
  • Basic notions of Fluxlang scripting including:
    • filtering by measurement, tags’ keys and values, fields’ keys and values
    • using aggregate windows for event summarization and time series synchronization
  • Knowledge about advanced data engineering tasks such as time series schemata exploration, and time series enrichment (would be helpful)
  • Knowledge about basic data science tasks such as trend detection, spread/correlation assessment, recognition of repetitive patterns (would be helpful)
  • We highly recommend completing the free self-paced Beginner Flux and Intermediate Flux courses if you have not attended an InfluxDays Flux training in the past.
  • All participants will need:
    • A laptop for class labs and exercises
    • A free account in InfluxDB Cloud (You will need your InfluxDB Cloud Org ID to complete registration)
    • An InfluxDB Community Slack account

“I really enjoyed the last Flux training by this team. I think the instructors Marco and Emanuele did a great job putting the right amount of information for the training. It was fun!”

Angelo Fausti, Software Engineer, Vera C. Rubin Observatory

Day One

Day Two

9:00 – 9:15 Setup and fast refresh of Flux (filtering, summarization and aggregate windows) 9:00 – 9:15 Day 1 Review and Q&A
9:15 – 10:00

Data Engineering with Advanced Flux – schema exploration

  • Case study
  • Explore measurements, tags and fields
  • Critical review of the modeling
9:15 – 10:00

Data Science with Advanced Flux – Correlation over time

  • The intuition
  • Building a parametric dashboard to correlate time series pairwise
  • Spurious correlation
  • Correlation is not causation
10:00 – 10:15 Q&A Break 10:00 – 10:15 Q&A Break
10:15 – 11:00

Data Engineering with Advanced Flux – Time series enrichment

  • The problem
  • The solution using join & map
  • The solution using pivoting & map
  • Join vs. pivoting
10:15 – 10:45

Data Science with Advanced Flux – Seasonality

  • Recognizing repetitive patterns
  • Advanced usages of graphs from InfluxDB Cloud UI to build a seasonal plot
11:00 – 11:15 Q&A Break 10:45  – 11:15 Q&A Break
11:15 – 12:00

Data Exploration with Advanced Flux – Trend

  • Trend detection methods
  • Smoothing techniques
  • MovingAverage
  • ExponentialMovingAverage
  • InfluxDB variable and parametric Flux Scripts
  • Detrending
11:15 – 11:45

Data Science with Advanced Flux – Seasonality

  • Deseasoning a time series to get the residual
  • Using histograms to access if a residual is good (a.k.a., stationary)
12:00 – 12:15

Q&A session

11:45 – 12:15

Q&A session

12:15 – 13:15 Lunch Break 12:15 – 13:15 Lunch Break
13:15 – 14:00

Data Exploration with Advanced Flux – Spread

  • The band graph
  • Showing how Max, mean, Min evolve over time with a Band graph
  • Mean +/- stddev Band graph
  • Mean vs Median
  • Skewness over time
  • Time series normalization
13:15 – 14:00

Data Science with Advanced Flux – forecasting

  • Assumptions
  • The forecasting process
  • Holt-Winters forecasting method
14:00 – 14:15 Q&A Break 14:00 – 14:15 Q&A Break
14:15 – 17:00

Bootcamp about time series exploration and office hours

14:15 – 17:00

Bootcamp about building and deploying real-time data analytics pipelines and office hours

17:00 – 18:00 Happy hour
Agenda and schedule are subject to change.

Day One

9:00 – 9:15

Setup and fast refresh of Flux (filtering, summarization and aggregate windows)

9:15 – 10:00

Data Engineering with Advanced Flux – schema exploration

  • Case study
  • Explore measurements, tags and fields
  • Critical review of the modeling
10:00 – 10:15 Q&A Break
10:15 – 11:00

Data Engineering with Advanced Flux – Time series enrichment

  • The problem
  • The solution using join & map
  • The solution using pivoting & map
  • Join vs. pivoting
11:00 – 11:15 Q&A Break
11:15 – 12:00

Data Exploration with Advanced Flux – Trend

  • Trend detection methods
  • Smoothing techniques
  • MovingAverage
  • ExponentialMovingAverage
  • InfluxDB variable and parametric Flux Scripts
  • Detrending
12:00 – 12:15

Q&A session

12:15 – 13:15 Lunch Break
13:15 – 14:00

Data Exploration with Advanced Flux – Spread

  • The band graph
  • Showing how Max, mean, Min evolve over time with a Band graph
  • Mean +/- stddev Band graph
  • Mean vs Median
  • Skewness over time
  • Time series normalization
14:00 – 14:15 Q&A Break
14:15 – 17:00

Bootcamp about time series exploration and office hours

Day Two

9:00 – 9:15

Day 1 Review and Q&A

9:15 – 10:00

Data Science with Advanced Flux – Correlation over time

  • The intuition
  • Building a parametric dashboard to correlate time series pairwise
  • Spurious correlation
  • Correlation is not causation
10:00 – 10:15 Q&A Break
10:15 – 10:45

Data Science with Advanced Flux – Seasonality

  • Recognizing repetitive patterns
  • Advanced usages of graphs from InfluxDB Cloud UI to build a seasonal plot
10:45  – 11:15 Q&A Break
11:15 – 11:45

Data Science with Advanced Flux – Seasonality

  • Deseasoning a time series to get the residual
  • Using histograms to access if a residual is good (a.k.a., stationary)
11:45 – 12:15 Q&A session
12:15 – 13:15 Lunch Break
13:15 – 14:00

Data Science with Advanced Flux – forecasting

  • Assumptions
  • The forecasting process
  • Holt-Winters forecasting method
14:00 – 14:15 Q&A Break
14:15 – 17:00

Bootcamp about building and deploying real-time data analytics pipelines and office hours

17:00 – 18:00 Happy hour

Agenda and schedule are subject to change.

Your Trainers

Emanuele-Della-Valle,-PhD

Emanuele Della Valle, PhD,

Associate Professor | Politecnico di Milano

Expert in semantic technologies and stream computing. Brander of stream reasoning: an approach to master the velocity and variety dimension of Big Data blending stream processing and AI +20 years experience in innovation and research projects.

Ignacio Van Droogenbroeck

Marco Balduini, PhD

Founder & CEO | Quantia Consulting

Expert in data processing, data integration, and data science technologies. Main contributor of the C-SPARQL Engine, author of Streaming Linked Data framework and FraPPE ontology ~10 years experience in innovation and research projects.

Riccardo Tommasini, PhD

Riccardo Tommasini, PhD

Associate Professor | The Institut National des Sciences Appliquées de Lyon

Riccardo Tommasini is an Associate Professor (Maître des Confèrenecs) at the Institut National des Sciences Appliquées de Lyon or INSA Lyon, France. Prior to join INSA Lyon, Riccardo was Assistant Professor of Data management at the University of Tartu, Estonia. Riccardo did his PhD at the Department of Electronics and Information of the Politecnico di Milano with a thesis on “Velocity on the Web”. The thesis investigates the velocity aspects that concern the Web environment, together with other challenges such as variety and volume. His research interests span Stream Processing, Knowledge Graphs, Logics and Programming Languages. Riccardo’s tutorial activities comprise Stream Reasoning Tutorials at ISWC 2017, ICWE 2018, ESWC 2019, and TheWebConf 2019, and DEBS 2019, IEEE Big Data 2021.

Hands-On Advanced Flux Training