Working with the data yourself

Load the free sample CSV, group it by token and venue, and write a finding you can defend.

Lesson 10 of 10 · 10 min read · All lessons

The free sample

The sample file holds real moves by tracked whale wallets over the last 7 days, for the 10 coins with the most tracked-whale volume, up to the latest 5,000 moves. It is refreshed weekly and licensed CC BY 4.0, so you can use it in class or in a project with the attribution “Deep Blue Alpha (deepbluealpha.io)”.

Line 1 is a comment; the header is line 2. Columns: time_utc, tx_hash, wallet, token, action, usd_value, venue_type, exchange.

A first analysis

Load it, then split by venue_type before adding anything up, so exchange transfers and DEX trades never get mixed:

import pandas as pd df = pd.read_csv("https://deepbluealpha.io/data/sample.csv", comment="#") flow = df.groupby(["token", "venue_type", "action"])["usd_value"].sum().unstack(fill_value=0) print(flow.round(0)) # Which wallets drive each token? print(df.groupby(["token", "wallet"])["usd_value"].sum().sort_values(ascending=False).head(10))

Write a defensible finding

A good finding names the window, the wallets (tracked whales), the kind of move, and its own doubt: “In the seven days to Sunday, tracked whales withdrew more ETH from exchanges than they deposited, mostly from three wallets.” Not: “Whales are buying ETH.”

Try it on a live page

  1. Download the sample (or load it with the code above).
  2. For one token, compute withdrawals minus deposits, using exchange rows only.
  3. Find how much of that figure came from the single largest wallet, then write one defensible sentence.
Open Free sample CSV →

Check yourself

1. Before summing the sample, you should split by…

Exchange transfers and DEX trades are different things.

2. The sample’s license lets you…

It is CC BY 4.0: attribution required.

3. Which finding is written honestly?

It names the group, the kind of move, the window and the concentration.

Go deeper

On-Chain Analysis for Beginners: How to Read Blockchain Data →How to Track Crypto Whales: The Complete Guide →
← 9. Measuring a signal honestly For instructors →

Describes past on-chain activity and how our data is built. Not investment advice. Method details: data methodology.