NFT Data Science pt. 1 - Intro to OpenSea API

By Alex Duffy elaborating on Adil Moujahid’s framework

Cryptocurrency is digital native currency, or currency made for and on a digital medium / the internet). In a similar vein, non-fungible tokens (NFT’s) are digital native THINGS that you can buy with said currency and own. THINGS that you hold in your secure wallet, THINGS can be read and interacted with by any website that you give permission to, THINGS that can often be bought sold and traded.

OpenSea is the first and largest NFT marketplace created in June 2018. During that month 26 NFT’s were sold for just under $$ 600 combined. Last month as of writing this (August 2021) 1.6 million NFT’s sold for $ 3.4 billion USD; more than all previous months combined. That may have been a fluke and NFT’s may be a fad, or it could be a great time to learn more, look at popular collections, and analyze their trends.

dune_OS.png Analytics dashboard on OpenSea made by user rchen8 - https://dune.xyz/rchen8/opensea

As an example I will be investigating the NFT collection Animetas, one that I am particularly interested in along with the Savage Droids collection which will be explored in part 2.

As you follow along, you can change the collection this notebook pulls data for. All you have to do is replace ‘asset_contract_address’ with the smart contract address of your choice. This address can be found through OpenSea

Note: You will have to change the start and and dates as well for data pulling. The initial query could take some time and require additional sleep time between loops to not throttle but once pulled and saved you can set dates to only query new data. That way you would only need to load saved data greatly reducing wait time for getting the most updated info.

droid.png

contract.png

Notebook Functionalities:

asset_contract_address = "0x18Df6C571F6fE9283B87f910E41dc5c8b77b7da5"
%matplotlib inline
from helpers import parse_events_data, parse_assets_data, parse_sale_data, parse_listing_data

import requests
import pandas as pd
from pandas_profiling import ProfileReport
import pickle

from statistics import *
import numpy as np
from scipy.stats import combine_pvalues

import matplotlib
import matplotlib.pyplot as plt
import seaborn as sns
plt.style.use('ggplot')

import time
from datetime import date, timedelta, datetime

import os
import glob

Simple demo of the OpenSea API

Shows all information possible from Assets and Events without pre-processing

Here we demonstrate how to requests information from the OpenSea API and what data comes in that response for Assets and Events.

  • Assets are the NFT’s themselves, e.g. Droid B-6763 pictured above
  • Events are transactions of NFT’s, e.g. sales, listings, bids entered, bids withdrawn, etc.

Assets

url = "https://api.opensea.io/api/v1/assets"

querystring = {"token_ids":list(range(0, 30)),
               "asset_contract_address":asset_contract_address,
               "order_direction":"desc",
               "offset":"0",
               "limit":"30"}
response = requests.request("GET", url, params=querystring)


# Getting asset data
assets = response.json()['assets']
if assets == []:
    print('empty')
sample_asset = pd.DataFrame.from_dict(assets[0], orient='index')
sample_asset
0
id 33619906
token_id 26
num_sales 0
background_color None
image_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_preview_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_thumbnail_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_original_url https://gateway.pinata.cloud/ipfs/QmbfuMdX9qiM...
animation_url None
animation_original_url None
name Animeta #26
description Animetas is a generative collection of 10101 u...
external_link None
asset_contract {'address': '0x18df6c571f6fe9283b87f910e41dc5c...
permalink https://opensea.io/assets/0x18df6c571f6fe9283b...
collection {'banner_image_url': 'https://lh3.googleuserco...
decimals 0
token_metadata https://ipfs.io/ipfs/QmQobwx36WXzNNWmFVvwNe4Pg...
owner {'user': {'username': 'joepayne'}, 'profile_im...
sell_orders None
creator {'user': {'username': 'Animetas_Vault'}, 'prof...
traits [{'trait_type': 'print', 'value': 'cat', 'disp...
last_sale None
top_bid None
listing_date None
is_presale False
transfer_fee_payment_token None
transfer_fee None

Events

url = "https://api.opensea.io/api/v1/events"

querystring = {"asset_contract_address":asset_contract_address,
               "event_type":"successful",
               "only_opensea":"true",
               "offset":0,
               "limit":"300"}
headers = {"Accept": "application/json"}

response = requests.request("GET", url, headers=headers, params=querystring)

# Getting event data
events = response.json()['asset_events']
if events == []:
    print('empty')
sample_event = pd.DataFrame.from_dict(assets[0], orient='index')
sample_event
0
id 33619906
token_id 26
num_sales 0
background_color None
image_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_preview_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_thumbnail_url https://lh3.googleusercontent.com/SEpYcwGMcQo6...
image_original_url https://gateway.pinata.cloud/ipfs/QmbfuMdX9qiM...
animation_url None
animation_original_url None
name Animeta #26
description Animetas is a generative collection of 10101 u...
external_link None
asset_contract {'address': '0x18df6c571f6fe9283b87f910e41dc5c...
permalink https://opensea.io/assets/0x18df6c571f6fe9283b...
collection {'banner_image_url': 'https://lh3.googleuserco...
decimals 0
token_metadata https://ipfs.io/ipfs/QmQobwx36WXzNNWmFVvwNe4Pg...
owner {'user': {'username': 'joepayne'}, 'profile_im...
sell_orders None
creator {'user': {'username': 'Animetas_Vault'}, 'prof...
traits [{'trait_type': 'print', 'value': 'cat', 'disp...
last_sale None
top_bid None
listing_date None
is_presale False
transfer_fee_payment_token None
transfer_fee None

Pull and Save NFT Data

Getting Assets Data

The code below collects assets data about all the assets. The API has a limit of 50 items per call and a max offset of 10,000. In order to circumvent this we increment token id instead of offest which leaves room for bugs if the tokens are not numbered in ascending order. We also need to be aware if the first token has the ID 0 or the ID of 1.

We use the parsing provided in Adil Moujahid’s framework to focus on these fields for each asset:

Asset Data Parsed:
  • ‘asset_id’
  • ‘creator_username’
  • ‘creator_address’
  • ‘owner_username’
  • ‘owner_address’
  • ‘traits’
  • ‘num_sales’

We also need to store this data somehow to be able to read from it locally instead of having to do the lengthy request every time we want to update our visuals.

One option is using pickle to save dataframes and lists. While convenient, it is not memory efficient. One option which we explore in part two is saving arrays as .npz files. Another alternative altogether suited for a more production ready environment would be feeding that data into a postgres SQL database.

def download_asset_info(save_location):
    if not os.path.isdir(save_location):
        os.makedirs(save_location)
    url = "https://api.opensea.io/api/v1/assets"
    listoassets = []

    for i in range(0, 3000):
        querystring = {"token_ids":list(range((i*30), (i*30)+30)),
                       "asset_contract_address":asset_contract_address,
                       "order_direction":"desc",
                       "offset":"0",
                       "limit":"30"}
        response = requests.request("GET", url, params=querystring)

        print(i, end=" ")
        if response.status_code != 200:
            print('error')
            print(response.json())
            break

        # Getting assets data
        assets = response.json()['assets']
        if assets == []:
            break
        # Parsing assets data
        parsed_assets = [parse_assets_data(asset) for asset in assets]
        # storing parsed events data into list
        listoassets.append(parsed_assets)
    
    # Flatten everything into one list
    listoassets = [item for sublist in listoassets for item in sublist]
    # Convert to df
    assets_df = pd.DataFrame(listoassets)
    
    with open(save_location + 'assets_df'+str(date.today())+r'.pkl', 'wb') as f:
        pickle.dump(assets_df, f)

Pull and Save NFT Sales

Getting sales transactions data

The code below collects all sale transactions data. The API has a limit of 300 items per call and 10,000 items total / response. Days with > 10,000 sales will require chunking (built in but you need to specify what hour chunks you want to chunk in) but most projects aren’t there yet. With that said in the NFT world there are often days where interest in a project can rapidly spike occasionally breaking that 10,000 item limit.

Sales Data Parsed
  • ‘is_bundle’
  • ‘event_id’
  • ‘seller_address’
  • ‘buyer_address’
  • ‘buyer_username’
  • ‘seller_username’
  • ‘timestamp’
  • ‘total_price’
  • ‘payment_token’
  • ‘usd_price’
  • ‘transaction_hash’
# Download sales info from start_date to end _date and save them all into their own day's files
# Default values 07/30/21 to today
def download_sales_info(save_location, start_date = date(2021, 7, 30), end_date = date.today()):
    if not os.path.isdir(save_location):
        os.makedirs(save_location)
    url = "https://api.opensea.io/api/v1/events"
    # get the number of days that we want to download and save sales for
    delta = end_date - start_date
    count_days = int(delta.days)
    
    for i in range(count_days+1):
        sales_that_day = []
        # set start and end of the day we are checking, if it's today set end to current time
        if date.today() == (start_date + timedelta(days=i)):
            before = datetime.now()
            after = datetime.combine((start_date + timedelta(days=i)), datetime.min.time())
        else:
            before = datetime.combine((start_date + timedelta(days=i+1)), datetime.min.time())
            after = datetime.combine((start_date + timedelta(days=i)), datetime.min.time())
        # There are too many transactions, now have to break them up by chunks in the day
        hour_chunks = 24
        chunk_count = 24/hour_chunks
        time.sleep(.5)

        for chunk in range(int(chunk_count)):
            end = False
            for j in range(0, 35):
                time.sleep(.5)
                # add the hour_chunk to the start of the day (after) time for each chunk
                # use the actual before if we pass it chronologically though
                changed_before = after + timedelta(hours=hour_chunks*(chunk+1)) - timedelta(minutes = 1)
                changed_after = after + timedelta(hours = hour_chunks*(chunk))
                
                # this should only happen on the last chunk of a split day or if on current day
                if before < changed_before:
                    changed_before = before
                    end = True

                querystring = {"asset_contract_address":asset_contract_address,
                               "event_type":"successful",
                               "only_opensea":"true",
                               "offset":j*300,
                               "occurred_before":changed_before,
                               "occurred_after":changed_after,
                               "limit":"300"}
                headers = {"Accept": "application/json"}

                response = requests.request("GET", url, headers=headers, params=querystring)


                print(j, end=" ")
                if response.status_code != 200:
                    print('error')
                    print(response.json())
                    break

                #Getting assets sales data
                event_sales = response.json()['asset_events']

                if event_sales == []:
                    end =True
                    break

                # Parsing asset sales data
                parsed_event_sales = [parse_sale_data(sale) for sale in event_sales]
                # storing parsed events data into list
                sales_that_day.append(parsed_event_sales)
                # check if the last date in the list is the same day as 
                last_date = (datetime.strptime(parsed_event_sales[0]['timestamp'], '%Y-%m-%dT%H:%M:%S'))
                print(last_date)
            if end:
                break
        sales_that_day = [item for sublist in sales_that_day for item in sublist]
        
        print(str(len(sales_that_day))+ " sales saved to" + save_location + "events_sales_list_" + str(start_date + timedelta(days=i))+'.pkl')
        with open(save_location + "events_sales_list_" + str(start_date + timedelta(days=i))+'.pkl', 'wb') as f:
            pickle.dump(sales_that_day, f)

Pull and Save NFT Listings

Getting assets listings transactions data

There are other events than just sales. There are listings, bids entered, bids withdrawn, listings cancelled, etc. To change which you are requesting you change the ‘event_type’ modifier. You can see the different event types and other information about the API here: https://docs.opensea.io/reference/retrieving-asset-events

Listing Data Parsed:
  • ‘is_bundle’
  • ‘event_id’
  • ‘seller_address’
  • ‘seller_username’
  • ‘created_date’
  • ‘starting_price’
  • ‘payment_token’
  • ‘usd_price’
# Download listings info from start_date to end _date and save them all into their own day's files
# Default values to first listing ever 7/30/21 to today
def download_listings_info(save_location, start_date = date(2021, 7, 30), end_date = date.today()):
    if not os.path.isdir(save_location):
        os.makedirs(save_location)
    url = "https://api.opensea.io/api/v1/events"
    # get the number of days that we want to download and save listings for
    delta = end_date - start_date
    count_days = int(delta.days)
    
    for i in range(count_days+1):
        listings_that_day = []
        # set start and end of the day we are checking, if it's today set end to current time
        if date.today() == (start_date + timedelta(days=i)):
            before = datetime.now()
            after = datetime.combine((start_date + timedelta(days=i)), datetime.min.time())
        else:
            before = datetime.combine((start_date + timedelta(days=i+1)), datetime.min.time())
            after = datetime.combine((start_date + timedelta(days=i)), datetime.min.time())
        # There are too many transactions, now have to break them up by chunks in the day
        hour_chunks = 24
        chunk_count = 24/hour_chunks
        time.sleep(.5)

        
        for chunk in range(int(chunk_count)):
            end = False
            for j in range(0, 35):
                time.sleep(.5)
                # add the hour_chunk to the start of the day (after) time for each chunk
                # use the actual before if we pass it chronologically though
                changed_before = after + timedelta(hours=hour_chunks*(chunk+1)) - timedelta(minutes = 1)
                changed_after = after + timedelta(hours = hour_chunks*(chunk))
                
                # this should only happen on the last chunk of a split day or if on current day
                if before < changed_before:
                    changed_before = before
                    end = True

                querystring = {"asset_contract_address":asset_contract_address,
                               "event_type":"created",
                               "only_opensea":"true",
                               "offset":j*300,
                               "occurred_before":changed_before,
                               "occurred_after":changed_after,
                               "limit":"300"}
                headers = {"Accept": "application/json"}

                response = requests.request("GET", url, headers=headers, params=querystring)


                print(j, end=" ")
                if response.status_code != 200:
                    print('error')
                    print(response.json())
                    break

                #Getting assets listings data
                event_listings = response.json()['asset_events']

                if event_listings == []:
                    end = True
                    break

                # Parsing events listings data
                parsed_event_listings = [parse_listing_data(listing) for listing in event_listings]
                # storing parsed events data into list
                listings_that_day.append(parsed_event_listings)
                # check if the last date in the list is the same day as 
                print(parsed_event_listings[0]['created_date'])
            if end:
                break
        listings_that_day = [item for sublist in listings_that_day for item in sublist]
        
        print(str(len(listings_that_day))+ " listings saved to" + save_location +
              "events_listings_list_" + str(start_date + timedelta(days=i))+'.pkl')
        with open(save_location + "events_listings_list_" + str(start_date + timedelta(days=i))+'.pkl', 'wb') as f:
            pickle.dump(listings_that_day, f)

Each of the download files saves a pickled list / DF of their information which can be loaded below for analysis

Sales and Listings downloads support start_date and end_date to avoid taking extra time to download data more than once! We print out information about how many assets and events are being downloaded. We also print any error messages in case there are errors with throttling or other limitations of the OpenSea API.

# Change location to suit your own needs
save_location = "./static/animetas/"

# assets
download_asset_info(save_location)

# SALES
# download sales info from start_date to end_date e.g. date(2021, 7, 30), date.today() - timedelta(days=1), etc.
# defaults to first day of sales to today

download_sales_info(save_location = save_location, start_date = date(2021,7,30))

# LISTINGS
# download listings info from start_date to end_date e.g. date(2021, 7, 30), date.today() - timedelta(days=1), etc.
# defaults to first day of listings to today

download_listings_info(save_location = save_location, start_date = date(2021,7,30))
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 0 2021-07-30 23:53:54
1 208 sales saved to./static/animetas/events_sales_list_2021-07-30.pkl
0 2021-07-31 23:58:48
1 2021-07-31 23:16:18
2 2021-07-31 22:46:43
3 2021-07-31 20:25:10
4 2021-07-31 17:57:16
5 2021-07-31 17:05:49
6 2021-07-31 16:40:37
7 2021-07-31 16:25:15
8 2021-07-31 00:07:41
9 2402 sales saved to./static/animetas/events_sales_list_2021-07-31.pkl
0 2021-08-01 23:58:21
1 2021-08-01 20:46:25
2 2021-08-01 19:19:50
3 2021-08-01 14:32:59
4 2021-08-01 07:49:15
5 2021-08-01 03:33:43
6 2021-08-01 01:15:38
7 2074 sales saved to./static/animetas/events_sales_list_2021-08-01.pkl
0 2021-08-02 23:58:37
1 2021-08-02 18:30:59
2 2021-08-02 12:34:09
3 2021-08-02 06:18:04
4 2021-08-02 00:55:51
5 1246 sales saved to./static/animetas/events_sales_list_2021-08-02.pkl
0 2021-08-03 23:58:03
1 2021-08-03 18:24:06
2 2021-08-03 14:42:46
3 2021-08-03 08:42:32
4 2021-08-03 03:00:39
5 1406 sales saved to./static/animetas/events_sales_list_2021-08-03.pkl
0 2021-08-04 23:57:53
1 2021-08-04 17:44:08
2 2021-08-04 11:30:46
3 2021-08-04 03:59:13
4 1009 sales saved to./static/animetas/events_sales_list_2021-08-04.pkl
0 2021-08-05 23:58:22
1 2021-08-05 15:22:18
2 2021-08-05 09:11:33
3 884 sales saved to./static/animetas/events_sales_list_2021-08-05.pkl
0 2021-08-06 23:48:51
1 2021-08-06 14:14:55
2 2021-08-06 09:19:54
3 2021-08-06 04:14:17
4 1035 sales saved to./static/animetas/events_sales_list_2021-08-06.pkl
0 2021-08-07 23:55:39
1 2021-08-07 07:45:34
2 411 sales saved to./static/animetas/events_sales_list_2021-08-07.pkl
0 2021-08-08 23:58:55
1 282 sales saved to./static/animetas/events_sales_list_2021-08-08.pkl
0 2021-08-09 23:51:26
1 163 sales saved to./static/animetas/events_sales_list_2021-08-09.pkl
0 2021-08-10 23:58:51
1 165 sales saved to./static/animetas/events_sales_list_2021-08-10.pkl
0 2021-08-11 23:53:26
1 180 sales saved to./static/animetas/events_sales_list_2021-08-11.pkl
0 2021-08-12 23:54:28
1 109 sales saved to./static/animetas/events_sales_list_2021-08-12.pkl
0 2021-08-13 23:58:20
1 219 sales saved to./static/animetas/events_sales_list_2021-08-13.pkl
0 2021-08-14 23:28:58
1 268 sales saved to./static/animetas/events_sales_list_2021-08-14.pkl
0 2021-08-15 23:39:20
1 154 sales saved to./static/animetas/events_sales_list_2021-08-15.pkl
0 2021-08-16 23:58:26
1 69 sales saved to./static/animetas/events_sales_list_2021-08-16.pkl
0 2021-08-17 23:56:11
1 65 sales saved to./static/animetas/events_sales_list_2021-08-17.pkl
0 2021-08-18 23:52:18
1 72 sales saved to./static/animetas/events_sales_list_2021-08-18.pkl
0 2021-08-19 23:34:56
1 56 sales saved to./static/animetas/events_sales_list_2021-08-19.pkl
0 2021-08-20 23:04:07
1 73 sales saved to./static/animetas/events_sales_list_2021-08-20.pkl
0 2021-08-21 23:50:54
1 72 sales saved to./static/animetas/events_sales_list_2021-08-21.pkl
0 2021-08-22 23:53:13
1 98 sales saved to./static/animetas/events_sales_list_2021-08-22.pkl
0 2021-08-23 23:52:43
1 167 sales saved to./static/animetas/events_sales_list_2021-08-23.pkl
0 2021-08-24 23:52:48
1 157 sales saved to./static/animetas/events_sales_list_2021-08-24.pkl
0 2021-08-25 23:53:35
1 155 sales saved to./static/animetas/events_sales_list_2021-08-25.pkl
0 2021-08-26 23:57:44
1 118 sales saved to./static/animetas/events_sales_list_2021-08-26.pkl
0 2021-08-27 22:40:35
1 85 sales saved to./static/animetas/events_sales_list_2021-08-27.pkl
0 2021-08-28 23:55:50
1 72 sales saved to./static/animetas/events_sales_list_2021-08-28.pkl
0 2021-08-29 23:54:15
1 108 sales saved to./static/animetas/events_sales_list_2021-08-29.pkl
0 2021-08-30 23:28:36
1 124 sales saved to./static/animetas/events_sales_list_2021-08-30.pkl
0 2021-08-31 23:37:32
1 108 sales saved to./static/animetas/events_sales_list_2021-08-31.pkl
0 2021-09-01 23:58:21
1 80 sales saved to./static/animetas/events_sales_list_2021-09-01.pkl
0 2021-09-02 23:34:16
1 49 sales saved to./static/animetas/events_sales_list_2021-09-02.pkl
0 2021-09-03 23:48:41
1 54 sales saved to./static/animetas/events_sales_list_2021-09-03.pkl
0 2021-09-04 23:06:16
1 63 sales saved to./static/animetas/events_sales_list_2021-09-04.pkl
0 2021-09-05 22:15:02
1 30 sales saved to./static/animetas/events_sales_list_2021-09-05.pkl
0 2021-09-06 23:10:12
1 34 sales saved to./static/animetas/events_sales_list_2021-09-06.pkl
0 2021-09-07 23:55:04
1 37 sales saved to./static/animetas/events_sales_list_2021-09-07.pkl
0 2021-09-08 23:40:49
1 33 sales saved to./static/animetas/events_sales_list_2021-09-08.pkl
0 2021-09-09 21:06:35
1 42 sales saved to./static/animetas/events_sales_list_2021-09-09.pkl
0 2021-09-10 23:38:58
1 36 sales saved to./static/animetas/events_sales_list_2021-09-10.pkl
0 2021-09-11 21:53:00
1 15 sales saved to./static/animetas/events_sales_list_2021-09-11.pkl
0 2021-09-12 21:45:47
1 15 sales saved to./static/animetas/events_sales_list_2021-09-12.pkl
0 2021-09-13 23:52:24
1 35 sales saved to./static/animetas/events_sales_list_2021-09-13.pkl
0 2021-09-14 21:53:57
1 29 sales saved to./static/animetas/events_sales_list_2021-09-14.pkl
0 2021-09-15 21:58:30
1 18 sales saved to./static/animetas/events_sales_list_2021-09-15.pkl
0 2021-09-16 23:00:38
1 33 sales saved to./static/animetas/events_sales_list_2021-09-16.pkl
0 2021-09-17 23:23:07
1 27 sales saved to./static/animetas/events_sales_list_2021-09-17.pkl
0 2021-09-18 18:05:11
1 16 sales saved to./static/animetas/events_sales_list_2021-09-18.pkl
0 2021-07-30T23:58:53.710235
1 2021-07-30T16:20:57.245742
2 585 listings saved to./static/animetas/events_listings_list_2021-07-30.pkl
0 2021-07-31T23:58:58.123650
1 2021-07-31T23:52:15.111110
2 2021-07-31T23:45:31.086338
3 2021-07-31T23:38:51.388399
4 2021-07-31T23:33:21.053337
5 2021-07-31T23:27:59.182295
6 2021-07-31T23:23:30.456417
7 2021-07-31T23:19:11.076039
8 2021-07-31T23:15:04.008444
9 2021-07-31T23:11:03.727700
10 2021-07-31T23:07:48.696061
11 2021-07-31T23:04:26.236027
12 2021-07-31T23:01:12.138839
13 2021-07-31T22:58:10.218460
14 2021-07-31T22:54:58.415967
15 2021-07-31T22:52:04.771594
16 2021-07-31T22:48:22.831806
17 2021-07-31T22:42:22.239504
18 2021-07-31T21:14:17.999282
19 2021-07-31T19:02:58.507187
20 2021-07-31T17:42:22.310564
21 2021-07-31T17:22:34.979755
22 2021-07-31T17:09:15.682858
23 2021-07-31T16:59:26.059875
24 2021-07-31T16:50:59.221043
25 2021-07-31T16:44:11.319561
26 2021-07-31T16:37:53.696755
27 2021-07-31T16:32:16.245629
28 2021-07-31T16:27:23.547086
29 2021-07-31T16:23:10.975968
30 2021-07-31T16:12:27.621686
31 2021-07-31T04:58:42.210409
32 9436 listings saved to./static/animetas/events_listings_list_2021-07-31.pkl
0 2021-08-01T23:58:50.305743
1 2021-08-01T23:21:35.013876
2 2021-08-01T22:48:32.943467
3 2021-08-01T22:24:00.345418
4 2021-08-01T22:04:00.853090
5 2021-08-01T21:38:51.654973
6 2021-08-01T21:20:36.189000
7 2021-08-01T21:04:12.956086
8 2021-08-01T20:50:56.648659
9 2021-08-01T20:40:41.392252
10 2021-08-01T20:30:09.315430
11 2021-08-01T20:21:57.012412
12 2021-08-01T20:14:40.136292
13 2021-08-01T20:05:16.663407
14 2021-08-01T19:44:56.662666
15 2021-08-01T19:21:20.748362
16 2021-08-01T18:04:01.728972
17 2021-08-01T16:48:03.486206
18 2021-08-01T15:13:22.231238
19 2021-08-01T14:03:06.852987
20 2021-08-01T12:21:24.120936
21 2021-08-01T10:05:00.540720
22 2021-08-01T08:22:23.372371
23 2021-08-01T06:54:23.347213
24 2021-08-01T05:21:04.958852
25 2021-08-01T04:28:24.534787
26 2021-08-01T03:38:19.522332
27 2021-08-01T02:53:08.176902
28 2021-08-01T02:23:35.317632
29 2021-08-01T01:55:56.193229
30 2021-08-01T01:34:22.094260
31 2021-08-01T01:20:25.640843
32 2021-08-01T01:06:50.860269
33 2021-08-01T00:57:24.496003
34 error
{'offset': ['ensure this value is less than or equal to 10000']}
10200 listings saved to./static/animetas/events_listings_list_2021-08-01.pkl
0 2021-08-02T23:58:42.158530
1 2021-08-02T21:27:05.420384
2 2021-08-02T19:50:19.762151
3 2021-08-02T18:30:09.155800
4 2021-08-02T17:22:34.693013
5 2021-08-02T16:25:05.955988
6 2021-08-02T15:58:07.233710
7 2021-08-02T14:39:28.940903
8 2021-08-02T12:54:02.516071
9 2021-08-02T09:48:40.848657
10 2021-08-02T07:20:17.930045
11 2021-08-02T05:35:43.911542
12 2021-08-02T04:21:24.841636
13 2021-08-02T02:42:34.893324
14 2021-08-02T02:07:52.024097
15 2021-08-02T01:50:21.694420
16 2021-08-02T00:57:06.832246
17 2021-08-02T00:33:39.622448
18 5293 listings saved to./static/animetas/events_listings_list_2021-08-02.pkl
0 2021-08-03T23:58:54.301149
1 2021-08-03T21:01:39.078805
2 2021-08-03T18:31:43.612613
3 2021-08-03T16:13:09.027712
4 2021-08-03T14:00:10.491476
5 2021-08-03T10:59:00.926665
6 2021-08-03T05:57:02.374349
7 2021-08-03T03:03:19.368042
8 2021-08-03T00:03:21.518569
9 2407 listings saved to./static/animetas/events_listings_list_2021-08-03.pkl
0 2021-08-04T23:56:14.439553
1 2021-08-04T19:49:59.779273
2 2021-08-04T16:21:12.987419
3 2021-08-04T11:47:18.994124
4 2021-08-04T05:54:42.866423
5 2021-08-04T01:40:10.264267
6 1648 listings saved to./static/animetas/events_listings_list_2021-08-04.pkl
0 2021-08-05T23:58:47.885848
1 2021-08-05T20:53:40.865251
2 2021-08-05T17:48:51.495635
3 2021-08-05T14:12:07.586983
4 2021-08-05T09:43:10.300419
5 2021-08-05T04:07:02.086689
6 1781 listings saved to./static/animetas/events_listings_list_2021-08-05.pkl
0 2021-08-06T23:58:28.883727
1 2021-08-06T21:13:54.612843
2 2021-08-06T19:01:01.215499
3 2021-08-06T16:47:51.890295
4 2021-08-06T14:58:51.220153
5 2021-08-06T13:09:36.987596
6 2021-08-06T09:25:20.660152
7 2021-08-06T06:11:15.830514
8 2021-08-06T02:23:32.144053
9 2538 listings saved to./static/animetas/events_listings_list_2021-08-06.pkl
0 2021-08-07T23:57:36.466500
1 2021-08-07T20:23:05.356304
2 2021-08-07T17:25:26.622379
3 2021-08-07T14:23:31.727188
4 2021-08-07T10:27:57.561503
5 2021-08-07T06:10:28.572554
6 2021-08-07T02:04:13.307321
7 1985 listings saved to./static/animetas/events_listings_list_2021-08-07.pkl
0 2021-08-08T23:58:50.159176
1 2021-08-08T20:25:02.093640
2 2021-08-08T17:58:28.474621
3 2021-08-08T15:17:12.409112
4 2021-08-08T12:46:30.279498
5 2021-08-08T09:02:30.351487
6 2021-08-08T05:26:07.575424
7 2021-08-08T02:00:18.315814
8 2316 listings saved to./static/animetas/events_listings_list_2021-08-08.pkl
0 2021-08-09T23:58:50.482126
1 2021-08-09T19:59:39.224037
2 2021-08-09T16:00:36.065315
3 2021-08-09T12:36:20.308764
4 2021-08-09T06:40:54.134525
5 2021-08-09T01:25:26.315291
6 1576 listings saved to./static/animetas/events_listings_list_2021-08-09.pkl
0 2021-08-10T23:57:50.557025
1 2021-08-10T18:50:51.131935
2 2021-08-10T12:33:41.747805
3 2021-08-10T06:22:22.467802
4 2021-08-10T02:18:59.002264
5 1330 listings saved to./static/animetas/events_listings_list_2021-08-10.pkl
0 2021-08-11T23:52:14.296555
1 2021-08-11T16:34:38.550153
2 2021-08-11T11:17:24.734275
3 2021-08-11T03:42:46.554752
4 1066 listings saved to./static/animetas/events_listings_list_2021-08-11.pkl
0 2021-08-12T23:47:39.194539
1 2021-08-12T16:37:29.639162
2 2021-08-12T06:05:39.545240
3 843 listings saved to./static/animetas/events_listings_list_2021-08-12.pkl
0 2021-08-13T23:57:23.698406
1 2021-08-13T14:06:24.284642
2 2021-08-13T04:53:05.889914
3 756 listings saved to./static/animetas/events_listings_list_2021-08-13.pkl
0 2021-08-14T23:52:01.670985
1 2021-08-14T14:39:59.312177
2 2021-08-14T03:55:13.947761
3 726 listings saved to./static/animetas/events_listings_list_2021-08-14.pkl
0 2021-08-15T23:58:04.352658
1 2021-08-15T16:51:43.224145
2 2021-08-15T07:25:07.464969
3 808 listings saved to./static/animetas/events_listings_list_2021-08-15.pkl
0 2021-08-16T23:58:11.603309
1 2021-08-16T17:19:58.912751
2 2021-08-16T09:25:52.803510
3 860 listings saved to./static/animetas/events_listings_list_2021-08-16.pkl
0 2021-08-17T23:55:03.501560
1 2021-08-17T14:46:30.401199
2 2021-08-17T04:51:16.046493
3 802 listings saved to./static/animetas/events_listings_list_2021-08-17.pkl
0 2021-08-18T23:58:44.122262
1 2021-08-18T14:08:22.149864
2 2021-08-18T03:39:28.780326
3 726 listings saved to./static/animetas/events_listings_list_2021-08-18.pkl
0 2021-08-19T23:56:24.757659
1 2021-08-19T13:11:59.893207
2 2021-08-19T01:27:40.072590
3 632 listings saved to./static/animetas/events_listings_list_2021-08-19.pkl
0 2021-08-20T23:58:00.381923
1 2021-08-20T06:28:44.353881
2 455 listings saved to./static/animetas/events_listings_list_2021-08-20.pkl
0 2021-08-21T23:53:58.687565
1 2021-08-21T07:10:44.315862
2 432 listings saved to./static/animetas/events_listings_list_2021-08-21.pkl
0 2021-08-22T23:58:35.177570
1 2021-08-22T10:02:34.396149
2 523 listings saved to./static/animetas/events_listings_list_2021-08-22.pkl
0 2021-08-23T23:58:25.754970
1 2021-08-23T06:02:21.482721
2 428 listings saved to./static/animetas/events_listings_list_2021-08-23.pkl
0 2021-08-24T23:56:29.802115
1 2021-08-24T10:29:42.322185
2 530 listings saved to./static/animetas/events_listings_list_2021-08-24.pkl
0 2021-08-25T23:57:50.163484
1 2021-08-25T07:55:43.003356
2 470 listings saved to./static/animetas/events_listings_list_2021-08-25.pkl
0 2021-08-26T23:53:12.726717
1 2021-08-26T04:24:19.733108
2 393 listings saved to./static/animetas/events_listings_list_2021-08-26.pkl
0 2021-08-27T23:53:46.220098
1 2021-08-27T07:04:55.500687
2 413 listings saved to./static/animetas/events_listings_list_2021-08-27.pkl
0 2021-08-28T23:27:07.140907
1 2021-08-28T09:18:06.172575
2 477 listings saved to./static/animetas/events_listings_list_2021-08-28.pkl
0 2021-08-29T23:55:52.589181
1 2021-08-29T13:32:48.757221
2 596 listings saved to./static/animetas/events_listings_list_2021-08-29.pkl
0 2021-08-30T23:55:30.100229
1 2021-08-30T15:44:20.657931
2 2021-08-30T05:47:58.678498
3 751 listings saved to./static/animetas/events_listings_list_2021-08-30.pkl
0 2021-08-31T23:51:19.390024
1 2021-08-31T13:49:53.130560
2 2021-08-31T04:03:15.795333
3 752 listings saved to./static/animetas/events_listings_list_2021-08-31.pkl
0 2021-09-01T23:53:50.450689
1 2021-09-01T16:40:55.657687
2 2021-09-01T08:39:31.383030
3 2021-09-01T00:39:47.906675
4 920 listings saved to./static/animetas/events_listings_list_2021-09-01.pkl
0 2021-09-02T23:54:30.573760
1 2021-09-02T12:26:20.065512
2 2021-09-02T02:32:06.589025
3 692 listings saved to./static/animetas/events_listings_list_2021-09-02.pkl
0 2021-09-03T23:58:48.160987
1 2021-09-03T09:52:28.190927
2 2021-09-03T01:56:50.889887
3 643 listings saved to./static/animetas/events_listings_list_2021-09-03.pkl
0 2021-09-04T23:39:47.583933
1 2021-09-04T04:44:24.516096
2 379 listings saved to./static/animetas/events_listings_list_2021-09-04.pkl
0 2021-09-05T23:58:49.575579
1 2021-09-05T02:36:51.392433
2 351 listings saved to./static/animetas/events_listings_list_2021-09-05.pkl
0 2021-09-06T23:57:53.514393
1 2021-09-06T12:54:03.150056
2 562 listings saved to./static/animetas/events_listings_list_2021-09-06.pkl
0 2021-09-07T23:53:26.767415
1 2021-09-07T13:20:44.728380
2 2021-09-07T02:40:46.617472
3 678 listings saved to./static/animetas/events_listings_list_2021-09-07.pkl
0 2021-09-08T23:56:06.112504
1 2021-09-08T09:49:36.542563
2 569 listings saved to./static/animetas/events_listings_list_2021-09-08.pkl
0 2021-09-09T23:58:56.151096
1 2021-09-09T11:30:41.484816
2 555 listings saved to./static/animetas/events_listings_list_2021-09-09.pkl
0 2021-09-10T23:53:45.904549
1 2021-09-10T08:48:10.506992
2 526 listings saved to./static/animetas/events_listings_list_2021-09-10.pkl
0 2021-09-11T23:57:17.806608
1 2021-09-11T09:24:14.329495
2 474 listings saved to./static/animetas/events_listings_list_2021-09-11.pkl
0 2021-09-12T23:52:39.595094
1 237 listings saved to./static/animetas/events_listings_list_2021-09-12.pkl
0 2021-09-13T23:56:17.731000
1 2021-09-13T10:45:24.864902
2 468 listings saved to./static/animetas/events_listings_list_2021-09-13.pkl
0 2021-09-14T23:56:37.352461
1 2021-09-14T08:41:36.166041
2 526 listings saved to./static/animetas/events_listings_list_2021-09-14.pkl
0 2021-09-15T23:57:31.530501
1 2021-09-15T08:26:09.723956
2 444 listings saved to./static/animetas/events_listings_list_2021-09-15.pkl
0 2021-09-16T23:56:51.635501
1 2021-09-16T12:57:39.008725
2 2021-09-16T01:12:03.934844
3 625 listings saved to./static/animetas/events_listings_list_2021-09-16.pkl
0 2021-09-17T23:57:19.471652
1 2021-09-17T12:17:59.555742
2 2021-09-17T01:37:06.696899
3 660 listings saved to./static/animetas/events_listings_list_2021-09-17.pkl
0 2021-09-18T19:45:39.521827
1 2021-09-18T07:04:49.655343
2 477 listings saved to./static/animetas/events_listings_list_2021-09-18.pkl

Load and Pre-Process Data

Functions for loading data from pickled files

# load the sales lists, combine them, and turn into a DF
def load_sales_info(save_location):
    files = [filename for filename in os.listdir(save_location) if filename.startswith('events_sales')]
    all_sales = []
    # load all files for sales by day
    for file in files:
        with open(str(save_location) + str(file), 'rb') as f:
            all_sales.append(pickle.load(f))
    
    # flatten into one list
    all_sales = [item for sublist in all_sales for item in sublist]
    # convert to dataframe
    events_sales_df = pd.DataFrame(all_sales)
    
    return events_sales_df

# load the listing lists, combine them, and turn into a DF
def load_listings_info(save_location):
    files = [filename for filename in os.listdir(save_location) if filename.startswith('events_listing')]
    all_listings = []
    # load all files for listings by day
    for file in files:
        with open(str(save_location) + str(file), 'rb') as f:
            all_listings.append(pickle.load(f))
    
    # flatten into one list
    all_listings = [item for sublist in all_listings for item in sublist]
    # convert to dataframe
    events_listings_df = pd.DataFrame(all_listings)
    
    return events_listings_df

# load most recent saved assets df
def load_assets_info(save_location):
    files = glob.glob(str(save_location)+'assets_df????-??-??.pkl')
    with open(max(files, key=os.path.getctime), 'rb') as f:
        return pickle.load(f)

Load and pre-process our saved files

Note: Currently dropping bundled items and those paid for with stable coins, only looking at ETH purchases

Here we load the data of our sales, listings, and assets in general. We also do preliminary transformations of the data from our sales and listings:

  • To make sure the data is uniform so we drop bundle transactions as well as those not paid in ETH
  • We drop any duplicates in case there were any accidentally included multiple times
  • Change the units of the ‘total_price’ column into ETH instead of ether - 1 ETH = 1*10^18 ether
  • Convert the ‘timestamp’ column into a datetime format to be able to compare and manipulate time
  • Create a total_price_USD column which is only a rough approximiation using the price of ETH in USD at the time of query
# load all our saved files

# SALES
events_sales_df = load_sales_info(save_location)
# Pre-processing
# Convert price from WEI to ETH & for now get rid of bundles and duplicates(?)
events_sales_df = events_sales_df[(events_sales_df['payment_token'] != 'USDC') & (events_sales_df['is_bundle'] == False)].copy()
events_sales_df = events_sales_df.loc[events_sales_df.astype(str).drop_duplicates().index]
events_sales_df['total_price'] = events_sales_df['total_price']/10.**18
# Change timestamp to datetime
events_sales_df['timestamp'] = pd.to_datetime(events_sales_df['timestamp'])
# Calculating the sale prices in USD
events_sales_df['total_price_usd'] = events_sales_df['total_price'] * events_sales_df['usd_price']


# LISTINGS
events_listings_df = load_listings_info(save_location)
# Pre-processing
# Convert price from WEI to ETH & for now get rid of bundles and duplicates(?)
events_listings_df = events_listings_df[(events_listings_df['payment_token'] != 'USDC') & (events_listings_df['is_bundle'] == False)].copy()
events_listings_df = events_listings_df.loc[events_listings_df.astype(str).drop_duplicates().index]
events_listings_df['starting_price'] = events_listings_df['starting_price']/10.**18
# Change timestamp to datetime
events_listings_df['created_date'] = pd.to_datetime(events_listings_df['created_date'])
# Calculating the sale prices in USD
events_listings_df['total_price_usd'] = events_listings_df['starting_price'] * events_listings_df['usd_price']


# assets
assets_df = load_assets_info(save_location)

Lets take a look at our assets and events

assets_df.head(5)
asset_id creator_username creator_address owner_username owner_address traits num_sales
0 26 Animetas_Vault 0x77c2783e24e397f14628b2ea56a6d967c62f9a36 joepayne 0x03f58f0cc44be4abc68b2df93c58514bb1196dc3 [{'trait_type': 'print', 'value': 'cat', 'disp... 0
1 23 Animetas_Vault 0x77c2783e24e397f14628b2ea56a6d967c62f9a36 Shakesbit2021 0x26a5a22570d8a4408cb21e333486ca04e60dec4d [{'trait_type': 'mask color', 'value': 'blue',... 1
2 25 Animetas_Vault 0x77c2783e24e397f14628b2ea56a6d967c62f9a36 leftwhaleshark 0x02f5d9c2b5376f8b9150cf148aa88a5fdd5dcc50 [{'trait_type': 'print', 'value': 'cat', 'disp... 0
3 24 Animetas_Vault 0x77c2783e24e397f14628b2ea56a6d967c62f9a36 EthRangers 0xda8a2e195b04f3dbd97c1a16dd8152634c951745 [{'trait_type': 'eye type', 'value': 'deep', '... 1
4 27 Animetas_Vault 0x77c2783e24e397f14628b2ea56a6d967c62f9a36 ET-ian 0xe62dd46747124854ad4c180386f22ea8b24e3673 [{'trait_type': 'mouth type', 'value': 'whimsi... 2
events_sales_df.tail(5)
is_bundle event_id seller_address buyer_address buyer_username seller_username timestamp total_price payment_token usd_price transaction_hash total_price_usd
14455 False 3922 0xd75eee9e0b62be2c0a0f673d655e0623a798397b 0x1d24c2beed8abd30626d6d2589f0e8b92ca8c2d2 Jay10369 cdl 2021-09-18 05:43:41 0.375 ETH 3419.49 0xee08a4880e442a63e4ad8572523a05eeffd3668e541a... 1282.30875
14456 False 8318 0x092f398f943cc1ff517f215215517d7f273f3ed9 0x69ab9f72e9e21af7604454f2824a4bcb0b7906c5 NFTholder333 JadeBasilisk222 2021-09-18 05:17:51 0.500 WETH 3412.11 0x2de90fc76e9b76e3acaaa45da3b01f9f882d5cb5cfda... 1706.05500
14457 False 7632 0x29ce78dc0f2a5b40dac18768c25f86e4199f3f1e 0xf7e886215cd82e9cacf1b1d2cd7ca5cc01d6488e 9825 Lucille-Ball3r 2021-09-18 05:03:12 0.400 ETH 3419.49 0x70bb7a46e8bc134164c8e34d3cdd8695da2a6ae25ce1... 1367.79600
14458 False 408 0x0d08ad2ab7893c04ecb460cbb6822b11c9e8904a 0x70732eb049383dd196faa3d89ff707e11b143f26 NFTTTT None 2021-09-18 03:42:39 0.350 ETH 3419.49 0x6a635140eec261e6be6a26bc6ce8f491e5075bd2dde6... 1196.82150
14459 False 4375 0x1a98347250498531758446ac22e605dceb46005c 0x9936845c322de64a068b145a993b18e3fdf60d36 None Altrine 2021-09-18 00:24:21 0.369 ETH 3419.49 0x2847b6378c68dcfbf28e70ff5198529f51e745148d76... 1261.79181

Basic Analysis

Analyzing events Data and events Sale Transactions

print("The database has information about %d assets." % len(assets_df))
print("The database has information about %d sale transactions." % len(events_sales_df))
print("The database has information about %d listing transactions." % len(events_listings_df))
The database has information about 10101 assets.
The database has information about 14408 sale transactions.
The database has information about 63884 listing transactions.

Pandas profiling is a great way to get detailed information on a dataset. It requires no work on our side but unfortunately takes a bunch of memory and is usually not able to be run on large datasets.

sales_report.png

# pandas profiling reports can be generated for smaller collections
"""
event_report = ProfileReport(assets_df, title="events Report", explorative=True)
event_sales_report = ProfileReport(events_sales_df, title="event Sales Report", explorative=True)
event_report.to_file("event_report.html")
event_sales_report.to_file("event_sales_report.html")
"""
'\nevent_report = ProfileReport(assets_df, title="events Report", explorative=True)\nevent_sales_report = ProfileReport(events_sales_df, title="event Sales Report", explorative=True)\nevent_report.to_file("event_report.html")\nevent_sales_report.to_file("event_sales_report.html")\n'

Analyzing events’ Sellers and Buyers

Understanding how many people are buying and selling a collection can give information on what kind of project it is. If there are many transactions all between not that many users, it could be a red flag and a reason to do more research into the project.

print("There are %d unique asset sellers." % len(events_sales_df['seller_address'].unique()))
print("There are %d unique asset buyers." % len(events_sales_df['buyer_address'].unique()))
There are 3636 unique asset sellers.
There are 4868 unique asset buyers.

Getting Top 10 events Buyers

The OpenSea API is convenient in that it allows developers to programatically access the username of buyers and sellers giving more context into WHO is doing the buying and selling.

buyers = []
for buyer_address in events_sales_df['buyer_address'].value_counts().index[:10]:
    buyer_data = {}
    buyer_data['buyer_address'] = buyer_address
    buyer_data['buyer_username'] = events_sales_df[events_sales_df['buyer_address'] == buyer_address]['buyer_username'].iloc[0]
    buyer_data['number_buys'] = len(events_sales_df[events_sales_df['buyer_address'] == buyer_address])
    buyer_data['min_price'] = events_sales_df[events_sales_df['buyer_address'] == buyer_address]['total_price'].min()
    buyer_data['max_price'] = events_sales_df[events_sales_df['buyer_address'] == buyer_address]['total_price'].max()
    buyer_data['mean_price'] = events_sales_df[events_sales_df['buyer_address'] == buyer_address]['total_price'].mean()
    buyers.append(buyer_data)
    
pd.DataFrame(buyers)
buyer_address buyer_username number_buys min_price max_price mean_price
0 0xa422bfff5daba6eeefaff84debf609edf0868c5f h49_vault 183 0.1300 1.00 0.661579
1 0xc6a7463a7ee700d035aff7bfd1ee198d680a4164 BobTheNailer 113 0.1190 2.75 0.393596
2 0x5d9049fccf5ae287ca2472713bb7fc6325dc7876 0X0923 110 0.2800 3.50 0.604512
3 0x7c8f4a31bfa6a2bc70e538dec4636da3c531abe8 NanoChip 98 0.1188 10.00 0.644079
4 0x0b2c327046b9f66e11752d1220bed0712b0d0188 likaboss 94 0.1590 0.80 0.299877
5 0x020ca66c30bec2c4fe3861a94e4db4a498a35872 MachiBigBrother 88 0.3190 7.50 1.827024
6 0x5cfa2c097d5f5ca13ff0ff3a2693cda8c026efd0 Animetas4Life 82 0.2000 0.25 0.224453
7 0x65217c49f9f1c4b5d468bcbf2942310cff530df1 C9F8EJJ8ED 76 0.1000 0.50 0.198117
8 0x27cfb1b71a3dcd6b7f19c273b70154a516b17c4f Sir_PantsALot 75 0.1000 9.00 0.803684
9 0x3ecc3c48310300eb71959ace8bd670960f0a8815 JiriK 62 0.2450 0.37 0.286806

Getting Top 10 events Sellers

sellers = []
for seller_address in events_sales_df['seller_address'].value_counts().index[:10]:
    seller_data = {}
    seller_data['seller_address'] = seller_address
    seller_data['seller_username'] = events_sales_df[events_sales_df['seller_address'] == seller_address]['seller_username'].iloc[0]
    seller_data['number_sales'] = len(events_sales_df[events_sales_df['seller_address'] == seller_address])
    seller_data['min_price'] = events_sales_df[events_sales_df['seller_address'] == seller_address]['total_price'].min()
    seller_data['max_price'] = events_sales_df[events_sales_df['seller_address'] == seller_address]['total_price'].max()
    seller_data['mean_price'] = events_sales_df[events_sales_df['seller_address'] == seller_address]['total_price'].mean()
    sellers.append(seller_data)
    
pd.DataFrame(sellers)
seller_address seller_username number_sales min_price max_price mean_price
0 0xbff79922fcbf93f9c30abb22322b271460c6bebb avarice 172 0.1800 1.850 0.299395
1 0x7c8f4a31bfa6a2bc70e538dec4636da3c531abe8 NanoChip 148 0.1500 1.400 0.404412
2 0x65217c49f9f1c4b5d468bcbf2942310cff530df1 C9F8EJJ8ED 101 0.2300 4.500 0.739970
3 0x88923378021bea85f9b09ce571a309e12c7d2262 8892 101 0.1200 2.750 0.361703
4 0xf889bdc0686274a419415298122b1086f0ce4a1d 0x125235 88 0.1500 0.180 0.163154
5 0x5cfa2c097d5f5ca13ff0ff3a2693cda8c026efd0 Animetas4Life 82 0.2900 3.000 0.584512
6 0x86494a5eb108ca2068a5bd55f617d430bc6d5eba 0xNew1 80 0.1189 3.490 0.256561
7 0x6708193ae7bbc4fc46b7e6a00af87a2b78fdb19f Flash_NFT 78 0.1800 1.999 0.320304
8 0x0b2c327046b9f66e11752d1220bed0712b0d0188 likaboss 76 0.0000 9.990 0.767936
9 0xf8a3db410668e79e1179fd54dfec9e78269694c0 None 69 0.1700 1.000 0.395754

Intersection of Top 10 Buyers and Top 10 sellers

top_10_buyers = events_sales_df['buyer_username'].value_counts().index[:10]
top_10_sellers = events_sales_df['seller_username'].value_counts().index[:10]
print(list(set(top_10_buyers) & set(top_10_sellers)))
['NanoChip', 'likaboss', 'Animetas4Life', 'C9F8EJJ8ED']

Getting Number of Sales between same Buyers and Sellers

(events_sales_df['seller_address'] + events_sales_df['buyer_address']).value_counts().value_counts()
1    13465
2      342
3       36
4       19
5        7
6        4
9        1
7        1
dtype: int64

Getting Top 10 asset Owners

We also want to see who is holding the most of any asset. This information is actually also available on the Etherscan website for any collection by looking at the token holders.

token.png

tokens.png

owners = []
for owner_address in assets_df['owner_address'].value_counts().index[:10]:
    owner_data = {}
    owner_data['owner_address'] = owner_address
    owner_data['owner_username'] = assets_df[assets_df['owner_address'] == owner_address]['owner_username'].iloc[0]
    owner_data['number_assets'] = len(assets_df[assets_df['owner_address'] == owner_address])
    owners.append(owner_data)

pd.DataFrame(owners)
owner_address owner_username number_assets
0 0xc6a7463a7ee700d035aff7bfd1ee198d680a4164 BobTheNailer 155
1 0x5d9049fccf5ae287ca2472713bb7fc6325dc7876 0X0923 119
2 0x8ee376de530fb9a734df676e7e4342b48355f483 DappPunk 96
3 0x04fc8679ed5475979f7d4930fedbd1c82e79db3a None 86
4 0x020ca66c30bec2c4fe3861a94e4db4a498a35872 MachiBigBrother 74
5 0xa8d3f65b6e2922fed1430b77ac2b557e1fa8da4a None 64
6 0x4d93c788b6e9771f1ee2f30242cd3892b631d8ed brent9two 48
7 0x4eafc1cbd90027d1723010f70d99873e4801f053 V1000 45
8 0x1d3643399e5534dd49f2b04f2f0615153bd209fd hive 43
9 0xb07c70eccb3373e9108a436cc1028d2ec6312ebf NFT11988 42
#### Getting total number of event Creators and Owners.
print("There are %d unique asset creators." % len(assets_df['creator_address'].unique()))
print("There are %d unique asset owners." % len(assets_df['owner_address'].unique()))
There are 2 unique asset creators.
There are 3918 unique asset owners.

Macro Visualization

Now that we have an understanding about the assets, no in order to get an understanding of the market dynamics to date, we can visualize our events.

  1. Total number of sales per Day (count & log count)
  2. Value of the day’s trades in ETH and USD equivalent
  3. The average and median price of all sales
  4. The floor price, or lowest price of an asset sold for during that period of time
  5. The highest price an asset sold for by day
  6. Total number of listings per day

These are interesting metrics that only begin to scratch the surface of a collection but can be the base of an expansive inquiry. There are many different views to work with and combining them into composite charts are often more informational. If you want to create production ready HTML visuals, I recommend exploring the Plotly package.

In the next part of this series, we will look at a different way of exploring a collection but this time digging one layer deeper and comparing the NFT’s by their often-unique traits.

events Sales Timelines

Total Number of Sales per Day

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').count()['total_price']
data = pd.DataFrame(data)
data.columns = ['Count assets Sold']
#ax = data.plot.bar(figsize=(18, 5))

plt.figure(figsize=(20,10))
ax = sns.barplot(x=data.index, y=data['Count assets Sold'], palette="winter")
plt.xlabel("Date")
plt.title("Sales Count")
ax.set_xticklabels(ax.get_xticklabels(),rotation = 30);
#ax.xaxis.set_major_locator(plt.MaxNLocator(100));

png

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').count()['total_price']
data = pd.DataFrame(data)
data.columns = ['Count assets Sold']
data['Count assets Sold'] = np.log(data['Count assets Sold'])
#ax = data.plot.bar(figsize=(18, 5))

plt.figure(figsize=(20,10))
ax = sns.barplot(x=data.index, y=data['Count assets Sold'], palette="winter")
plt.xlabel("Date")
plt.ylabel("Count assets Sold (LOG values)")
plt.title("Sales Count (LOG)")
ax.set_xticklabels(ax.get_xticklabels(),rotation = 30);
#ax.xaxis.set_major_locator(plt.MaxNLocator(100));

png

Total Sales per Day in ETH

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').sum()['total_price']
data = pd.DataFrame(data)
data.columns = ['total_price']

plt.figure(figsize=(20,10))
ax=sns.barplot(x=data.index, y=data['total_price'], palette="winter")
plt.xlabel("Date")
plt.title("Sales Count")

ax.set_title("Timeline of Total asset Sales in ETH", fontsize=18)
ax.set_ylabel("Sales in ETH", fontsize=18);
ax.set_xticklabels(ax.get_xticklabels(),rotation = 30);

png

Total Sales per day in USD

data = events_sales_df[['timestamp', 'total_price_usd']].resample('D', on='timestamp').sum()['total_price_usd']
ax = data.plot(figsize=(18,6), color="purple", linewidth=1, marker='o', markerfacecolor='grey', markeredgewidth=0)

ax.set_alpha(0.8)
ax.set_title("Timeline of Total asset Sales in Million USD", fontsize=18)
ax.set_ylabel("Sales in Million USD", fontsize=18);

dates = list(data.index)
values = list(data.values)

for i, j in zip(dates, values):
    ax.annotate(s="{:.2f}".format(j/10.**6), xy=(i, j), rotation=45)
<ipython-input-28-9e70797a0a1c>:12: MatplotlibDeprecationWarning: The 's' parameter of annotate() has been renamed 'text' since Matplotlib 3.3; support for the old name will be dropped two minor releases later.
  ax.annotate(s="{:.2f}".format(j/10.**6), xy=(i, j), rotation=45)

png

Asset Prices Timelines

Average asset Price per Day in ETH

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').mean()['total_price']
ax = data.plot(figsize=(18,6), color="green", linewidth=1, marker='o', markerfacecolor='grey', markeredgewidth=0)

ax.set_alpha(0.8)
ax.set_title("Timeline of Average asset Price in ETH", fontsize=18)
ax.set_ylabel("Average Price in ETH", fontsize=18);

dates = list(data.index)
values = list(data.values)

png

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').median()['total_price']
ax = data.plot(figsize=(18,6), color="green", linewidth=1, marker='o', markerfacecolor='grey', markeredgewidth=0)

ax.set_alpha(0.8)
ax.set_title("Timeline of Median asset Price in ETH", fontsize=18)
ax.set_ylabel("Median Price in ETH", fontsize=18);
#ax.annotate(s='sdsdsds', xy=(1, 1))

dates = list(data.index)
values = list(data.values)

png

Floor asset Price per Day in ETH

data = events_sales_df[['timestamp', 'total_price']].resample('240min', on='timestamp').min()['total_price']
ax = data.plot(figsize=(18,6), color="orange", linewidth=1, marker='o', markerfacecolor='grey', markeredgewidth=0)

ax.set_alpha(0.8)
ax.set_title("Timeline of Floor asset Price in ETH", fontsize=18)
ax.set_ylabel("Floor Price in ETH", fontsize=18);

dates = list(data.index)
values = list(data.values)

for idx, (d, v) in enumerate(zip(dates, values)):
    if idx%5 == 0:
        ax.annotate(s="{:.5f}".format(v), xy=(d, v), rotation=45)
<ipython-input-37-fb01d7c24f1e>:13: MatplotlibDeprecationWarning: The 's' parameter of annotate() has been renamed 'text' since Matplotlib 3.3; support for the old name will be dropped two minor releases later.
  ax.annotate(s="{:.5f}".format(v), xy=(d, v), rotation=45)

png

Max event Price per Day in ETH

data = events_sales_df[['timestamp', 'total_price']].resample('D', on='timestamp').max()['total_price']
ax = data.plot(figsize=(18,6), color="red", linewidth=1, marker='o', markerfacecolor='grey', markeredgewidth=0)

ax.set_alpha(0.8)
ax.set_title("Timeline of Max asset Price in ETH", fontsize=18)
ax.set_ylabel("Max Price in ETH", fontsize=18);

dates = list(data.index)
values = list(data.values)

for i, j in zip(dates, values):
    ax.annotate(s="{:.0f}".format(j), xy=(i, j+1), rotation=45)
<ipython-input-32-cb48f2fde8ca>:12: MatplotlibDeprecationWarning: The 's' parameter of annotate() has been renamed 'text' since Matplotlib 3.3; support for the old name will be dropped two minor releases later.
  ax.annotate(s="{:.0f}".format(j), xy=(i, j+1), rotation=45)

png

Assets Listings Timelines

Total Number of listings per Day

data = events_listings_df[['created_date', 'starting_price']].resample('D', on='created_date').count()['starting_price']
data = pd.DataFrame(data)
data.columns = ['Count assets Listed']
#ax = data.plot.bar(figsize=(18, 5))

plt.figure(figsize=(20,10))
ax = sns.barplot(x=data.index, y=data['Count assets Listed'], palette="winter")
plt.xlabel("Date")
plt.title("listings Count")
ax.set_xticklabels(ax.get_xticklabels(),rotation = 30);

png

data
Count assets Listed
created_date
2021-07-30 585
2021-07-31 9198
2021-08-01 10139
2021-08-02 5282
2021-08-03 2395
2021-08-04 1640
2021-08-05 1774
2021-08-06 2530
2021-08-07 1981
2021-08-08 2306
2021-08-09 1568
2021-08-10 1323
2021-08-11 1062
2021-08-12 836
2021-08-13 748
2021-08-14 721
2021-08-15 807
2021-08-16 859
2021-08-17 796
2021-08-18 722
2021-08-19 631
2021-08-20 453
2021-08-21 432
2021-08-22 519
2021-08-23 427
2021-08-24 530
2021-08-25 468
2021-08-26 393
2021-08-27 412
2021-08-28 477
2021-08-29 596
2021-08-30 751
2021-08-31 750
2021-09-01 919
2021-09-02 689
2021-09-03 642
2021-09-04 379
2021-09-05 350
2021-09-06 561
2021-09-07 677
2021-09-08 569
2021-09-09 555
2021-09-10 526
2021-09-11 474
2021-09-12 235
2021-09-13 468
2021-09-14 526
2021-09-15 444
2021-09-16 625
2021-09-17 657
2021-09-18 477

References

[0] Data mining Meebits

[1] Fungibility - Wikipedia

[2] A Practical Introduction to NFTs using Solidity and Legos

[3] Counterparty - Wikipedia

[4] Counterparty - Bitcoinwiki

[5] Rare Pepe Gets Blockchained, Made Into Tradable Counterparty Tokens


<
Blog Archive
Archive of all previous blog posts
>
Blog Archive
Archive of all previous blog posts