Update on Aave Analytics and Growth Dashboard - Omni Analytics

Hello Aave Community,

We’re the Omni Analytics team and we wanted to give an update on the progress we’ve made on our Aave Grant, the Aave Analytics and Growth Dashboard.

Current Progress

  1. We have optimized nearly every query in the existing Aave Growth Dashboard on Dune Analytics, in many cases improving the runtime by a factor of 100. This was done by performing more intelligent joins with the USD pricing data for the various tokens. The upshot of this is that nearly all queries run in a matter of seconds, and queries which were previously hitting Dune query limits on occasion are now executing successfully. Dune Analytics

  2. We created an initial version of the Aave Growth Dashboard on Polygon. We created Polygon versions of each query, and queries which do not depend on USD pricing data have been included in the Dashboard. Dune Analytics

  3. We’ve interfaced with the Covalent API in order to reconstruct approximately 2/3 of the existing queries from the Dune Analytics dashboard, using R and the tidyverse suite of packages. The visualizations have been written in ggplot2, allowing for significant refinement and flexibility, as well as aesthetic improvements, compared to the Dune dashboard.

  4. The Aave data from the Covalent API has been structured in raw and aggregated formats so that they can be served up and provided to the community for further analysis. We have begun scoping out the initial API which will allow access to this structured Aave data.

Next Steps

  1. The Polygon dashboard requires USD pricing data in order for many of the volume metric-based queries to be fully ported and visible. We are working with the team at Dune to get this data added, and will update the Dashboard accordingly once it is ready.

  2. We are continuing to validate the Aave growth metrics obtained via the Covalent API to ensure they are consistent with the previously validated Dune metrics.

  3. We will begin constructing the RShiny-based dashboard containing the available ported queries, and in the process optimize elements of the visuals, including improved axis labels, scaling, color schemes, and other visual elements of the plots.

  4. The remaining queries, particularly on Economics and Staking, will be ported to R and use data derived from the Covalent API. Once ready, the plots for these will be reconstructed in R as well.

Screenshots



If there are any ideas for additional analyses, or improvements to the existing analysis from the Dune dashboard, we would welcome any and all suggestions. A large motivation behind this project was to take advantage of the significant additional functionality afforded to us by the R ecosystem. As the existing results are ported, we can begin to take advantage of this to calculate statistics that would not be possible inside of Dune.

We hope you enjoy the progress!

4 Likes

Terrific progress Eric and the Omni team. It’s been great working with you all via Aave Grants and hope that this is the start of a long relationship with the Aave ecosystem!

2 Likes

Hello Aave Community!

Omni Analytics here again with another update for you on the Aave Analytics and Growth Dashboard. We’ve made some significant progress that we wanted to share:

  1. We have a first version of the dashboard, including most of the metrics we’ve computed thus far, available to play with. We’ve written the dashboard in RShiny, with Flexdashboard powering the layout. Flexdashboard provides an extremely customizable but still easy-to-use layout engine, powered by RMarkdown. If you’re interested in some more details about Flexdashboard, please see the docs here: R Markdown Format for Flexible Dashboards • flexdashboard But at a high level, we now have:
    a. The Covalent API for data access
    b. The R engine for data parsing, cleaning, and manipulation
    c. The Shiny server for dynamic and interactive computations
    d. The Flexdashboard engine for templating and UI design of the dashboard
    We think this stack provides a significant amount of flexibility both in terms of backend and frontend development, and future extensions.

  2. We have cleaned up much of the analysis and posted initial versions of the code for both the backend API calls and the dashboard itself on our GitHub repository here: GitHub - Omni-Analytics-Group/Aave-Growth-Dashboard: Our R-based dashboard on the growth of Aave

Some next steps we will be taking:

  1. Port the remaining queries from the Dune dashboard
  2. Include them in the RShiny dashboard
  3. Clean up and upload the remaining analysis code to the GitHub
  4. Further refine aesthetic elements of the dashboard and the visuals within the dashboard to better highlight the most pertinent Aave growth data.
  5. Setup a hosting solution for the web-accessible version of the dashboard

We hope you like the update, and we will follow up soon when there is more to share! Thanks for all the support up to this point.


2 Likes

Awesome, thanks for another update Eric! I shared on the AGD Twitter to broadcast to the wider ecosystem

1 Like

Hello Aave Community!

Omni Analytics is back with another update on our grant. Since last time, we’ve focused on two areas:

  1. Aave on Polygon Dune Dashboard - The major updates here are (1) the introduction of several new outputs / queries, (2) the refinement of existing queries. We switched from using the Dune User Generated pricing data, to the official prices.usd table which is now available. We updated nearly every query to make this change, and did some cross-checking of the outputs. This should now be the most accurate picture yet of the growth of Aave on Polygon! Please let us know if you have any comments.

  2. Aave on R Dashboard. We have added more options for interaction with the Aave on R dashboard. Some new features include:

  • The ability to select a date range for every plot
  • Aave-style branding on the plots
  • New transactions and user metrics
  • Updated design and layout


Next steps:

  1. Finalize and include missing Polygon and R dashboard queries from the Aave Dune Mainnet dashboard.
  2. Add more interactivity and output results in the R dashboard

As always, we welcome your comments and feedback!

2 Likes