Learning about Recommendation Systems with Anime Data

Using the Anime Recommendation Dataset from Kaggle, I explore the different types of recommendation systems and how they work.
Machine Learning
Ranking
Recommender Systems
Author

Daniel Fat

Published

February 12, 2020

Content-Based Filtering

Data Loading

imports

we’ve got some basic system imports, 🐼 and num🐍, 📊lib and fancy SimpleNamespace for using dictionaries cuz '"[] are annoying

Code 
import os,re,zipfile
import pandas as pd
import numpy as np
from types import SimpleNamespace
from matplotlib import pyplot as plt
import itertools
plt.style.use('seaborn')
base_size = 10
sizes = SimpleNamespace(**dict(small=(1*base_size,1*base_size),medium=(2*base_size,2*base_size),large=(3*base_size,3*base_size)))

Downlaod the data

Code 
%%time
zip_name = 'anime-recommendations-database.zip'
if not os.path.exists(zip_name):
  os.environ['KAGGLE_USERNAME'] = "" # username from the json file
  os.environ['KAGGLE_KEY'] = "" # key from the json file
  !kaggle datasets download CooperUnion/anime-recommendations-database
Downloading anime-recommendations-database.zip to /content
100% 25.0M/25.0M [00:00<00:00, 38.8MB/s]
100% 25.0M/25.0M [00:00<00:00, 83.3MB/s]
CPU times: user 24.8 ms, sys: 4.82 ms, total: 29.6 ms
Wall time: 2.41 s

Unzip the data

Code 
with zipfile.ZipFile(zip_name, 'r') as zip_ref:
    zip_ref.extractall(zip_name.split('.')[0])

Let’s find the files

Code 
os.listdir(zip_name.split('.')[0])
['anime.csv', 'rating.csv']

Have a look throught the dataset

Code 
anime = pd.read_csv(zip_name.split('.')[0] + '/anime.csv')
anime.genre = anime.genre.fillna('None')
anime
anime_id name genre type episodes rating members
0 32281 Kimi no Na wa. Drama, Romance, School, Supernatural Movie 1 9.37 200630
1 5114 Fullmetal Alchemist: Brotherhood Action, Adventure, Drama, Fantasy, Magic, Mili... TV 64 9.26 793665
2 28977 Gintama° Action, Comedy, Historical, Parody, Samurai, S... TV 51 9.25 114262
3 9253 Steins;Gate Sci-Fi, Thriller TV 24 9.17 673572
4 9969 Gintama&#039; Action, Comedy, Historical, Parody, Samurai, S... TV 51 9.16 151266
... ... ... ... ... ... ... ...
12289 9316 Toushindai My Lover: Minami tai Mecha-Minami Hentai OVA 1 4.15 211
12290 5543 Under World Hentai OVA 1 4.28 183
12291 5621 Violence Gekiga David no Hoshi Hentai OVA 4 4.88 219
12292 6133 Violence Gekiga Shin David no Hoshi: Inma Dens... Hentai OVA 1 4.98 175
12293 26081 Yasuji no Pornorama: Yacchimae!! Hentai Movie 1 5.46 142

12294 rows × 7 columns

Data Aggregation

Let’s get a list of unique genre

Code 
unique_genres = list(set([x.strip() for x in list(itertools.chain(*anime.genre.fillna('None').str.split(',')))]))
for gen in unique_genres:
  anime[gen] = 0

For each row we create something similar with the output of pd.get_dummies where we check the found genres against all the unique ones resulting into an array of 1 and 0 where given genres vs unique_genres

Code 
def binary_match(x,y):
  x = [t.strip() for t in x]
  y = [t.strip() for t in y]
  matches = dict(zip(y,np.zeros(len(y))))
  for j in y:
    if j in x:
      matches[j] = 1
    else:
      matches[j] = 0
  return matches
Code 
test = anime.genre.apply(lambda x: binary_match(x.split(','),unique_genres))
Code 
binary_df = pd.DataFrame(test.values.tolist())
Code 
anime[unique_genres] = binary_df[unique_genres]
Code 
anime
anime_id name genre type episodes rating members Military Harem Shoujo Ai Hentai Action Space Ecchi Magic Josei Yaoi Cars Shounen Ai Police Kids Seinen Psychological Drama Music Shoujo School None Fantasy Shounen Vampire Sci-Fi Super Power Horror Demons Sports Supernatural Slice of Life Parody Dementia Romance Yuri Game Adventure Historical Samurai Thriller Martial Arts Comedy Mecha Mystery
0 32281 Kimi no Na wa. Drama, Romance, School, Supernatural Movie 1 9.37 200630 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0
1 5114 Fullmetal Alchemist: Brotherhood Action, Adventure, Drama, Fantasy, Magic, Mili... TV 64 9.26 793665 1 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
2 28977 Gintama° Action, Comedy, Historical, Parody, Samurai, S... TV 51 9.25 114262 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 0 0 1 0 0
3 9253 Steins;Gate Sci-Fi, Thriller TV 24 9.17 673572 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
4 9969 Gintama&#039; Action, Comedy, Historical, Parody, Samurai, S... TV 51 9.16 151266 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 0 0 1 0 0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
12289 9316 Toushindai My Lover: Minami tai Mecha-Minami Hentai OVA 1 4.15 211 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
12290 5543 Under World Hentai OVA 1 4.28 183 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
12291 5621 Violence Gekiga David no Hoshi Hentai OVA 4 4.88 219 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
12292 6133 Violence Gekiga Shin David no Hoshi: Inma Dens... Hentai OVA 1 4.98 175 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
12293 26081 Yasuji no Pornorama: Yacchimae!! Hentai Movie 1 5.46 142 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

12294 rows × 51 columns

Looks good, right ?

KMeans Clustering for groups

Code 
from sklearn.cluster import KMeans

We use the number of unique genres as the number of clusters for a fast testing

Code 
kmeans = KMeans(n_clusters=len(unique_genres), random_state=0).fit(anime[unique_genres])
anime['clusters'] = kmeans.labels_
Code 
anime.clusters.value_counts().plot.barh(figsize=(10,10))

Then we define a searching by term function getting all the animes from that cluster sorted by rating

Code 
def awesome_find(df,search_term):
  tmpdf = df.copy()
  for x in search_term.split():
    tmpdf = tmpdf[tmpdf.name.str.upper().str.contains(x.strip().upper())]
  return df[df.clusters == tmpdf.sort_values(by='rating',ascending=False)[:1].clusters.values[0]][['anime_id','name','genre','type','episodes','members']]

Code 
awesome_find(anime,'Akame ga Kill')
anime_id name genre type episodes members
1 5114 Fullmetal Alchemist: Brotherhood Action, Adventure, Drama, Fantasy, Magic, Mili... TV 64 793665
24 164 Mononoke Hime Action, Adventure, Fantasy Movie 1 339556
101 18115 Magi: The Kingdom of Magic Action, Adventure, Fantasy, Magic, Shounen TV 25 245026
121 486 Kino no Tabi: The Beautiful World Action, Adventure, Slice of Life TV 13 102822
133 2685 Tsubasa: Tokyo Revelations Action, Adventure, Drama, Fantasy, Romance, Sh... OVA 3 57963
... ... ... ... ... ... ...
11076 30736 Shingeki no Bahamut: Virgin Soul Action, Adventure, Demons, Fantasy, Magic, Sup... TV 24 20953
11085 34086 Tales of Zestiria the X (2017) Action, Adventure, Fantasy TV 12 10848
11644 1331 Dragon Pink Action, Adventure, Comedy, Fantasy, Hentai, Ma... OVA 3 3467
11871 2185 Words Worth Gaiden Action, Adventure, Fantasy, Hentai OVA 2 2402
11957 1763 Midnight Panther Action, Adventure, Fantasy, Hentai OVA 2 1144

211 rows × 6 columns

Code 
from wordcloud import WordCloud
Code 
def create_wc(list_of_lists):
  words = list()
  for x in list_of_lists:
    for y in x.split(','):
      words.append(y.strip())
      
  # Create and generate a word cloud image:
  wordcloud = WordCloud(max_font_size=50, max_words=100, background_color="white").generate(' '.join(words))
  return wordcloud
Code 
def plot_figures(figures, nrows = 1, ncols=1):
    fig, axeslist = plt.subplots(ncols=ncols, nrows=nrows,figsize=(20,20))
    for ind,title in zip(range(len(figures)), figures):
        axeslist.ravel()[ind].imshow(figures[title], cmap=plt.jet())
        axeslist.ravel()[ind].set_title(f'Most Freqent words for the group {title+1}')
        axeslist.ravel()[ind].set_axis_off()
    plt.tight_layout() # optional
Code 
clouds = dict()
for x in anime.clusters.unique():
  clouds[x] = create_wc(anime[anime.clusters == x].genre.values.tolist())

Let’s see our clusters

Code 
plot_figures(clouds,9,5)

Cosine Similarity

Code 
from sklearn.metrics.pairwise import cosine_similarity

Our function computes the cosine similarity between the genre_vector of the given search term against all of the others

Code 
def search_cosine(df,search):
  tmpdf = df.copy()
  search_term = tmpdf[tmpdf.name == search][unique_genres].values
  tmpdf['similarity'] = cosine_similarity(search_term,tmpdf[unique_genres].values)[0].tolist()

  return tmpdf.sort_values(by='similarity',ascending=False)[['anime_id','name','genre','type','episodes','members','similarity']]

The only do downgrade on this approach is that the search has to be an exact match, however this can be solved through an elasticsearch index

Code 
src = search_cosine(anime,'Sword Art Online')
src
anime_id name genre type episodes members similarity
2132 21881 Sword Art Online II Action, Adventure, Fantasy, Game, Romance TV 24 537892 1.000000
804 11757 Sword Art Online Action, Adventure, Fantasy, Game, Romance TV 25 893100 1.000000
3392 20021 Sword Art Online: Extra Edition Action, Adventure, Fantasy, Game, Romance Special 1 121722 1.000000
10908 31765 Sword Art Online Movie: Ordinal Scale Action, Adventure, Fantasy, Game, Romance Movie 1 50944 1.000000
5078 27891 Sword Art Online II: Debriefing Action, Adventure, Fantasy, Game Special 1 39092 0.894427
... ... ... ... ... ... ... ...
7737 29944 Kuchao Dementia Movie 1 184 0.000000
7736 21335 Double Circle Sci-Fi, Super Power ONA 6 4249 0.000000
7735 32765 Cherry Blossom (Music) Music Music 1 113 0.000000
2255 10232 Tamayura: Hitotose Comedy, Drama, Slice of Life TV 12 25760 0.000000
12293 26081 Yasuji no Pornorama: Yacchimae!! Hentai Movie 1 142 0.000000

12294 rows × 7 columns

Collaborative Filtering

Code 
rating = pd.read_csv(zip_name.split('.')[0]+'/rating.csv')
rating
user_id anime_id rating
0 1 20 -1
1 1 24 -1
2 1 79 -1
3 1 226 -1
4 1 241 -1
... ... ... ...
7813732 73515 16512 7
7813733 73515 17187 9
7813734 73515 22145 10
7813735 73516 790 9
7813736 73516 8074 9

7813737 rows × 3 columns

Code 
mrating = rating.merge(rating.groupby('user_id').agg({'rating':'mean'}).reset_index(),on='user_id')
mrating = mrating.drop(mrating[mrating.rating_x < mrating.rating_y].index)
mrating
user_id anime_id rating_x rating_y
47 1 8074 10 -0.712418
81 1 11617 10 -0.712418
83 1 11757 10 -0.712418
101 1 15451 10 -0.712418
153 2 11771 10 2.666667
... ... ... ... ...
7813730 73515 13659 8 7.719388
7813733 73515 17187 9 7.719388
7813734 73515 22145 10 7.719388
7813735 73516 790 9 9.000000
7813736 73516 8074 9 9.000000

4262566 rows × 4 columns

PCA

Code 
from sklearn.decomposition import PCA
Code 
n_comps = 3
Code 
comp = pd.DataFrame(PCA(n_components=n_comps).fit_transform(mrating[['user_id','anime_id','rating_x']]),columns=[f'comp_{x}'for x in range(1,4)])
comp['anime_id'] = mrating['anime_id']
comp = comp.merge(anime[['anime_id','name']],right_on='anime_id',left_on='anime_id',how='inner')
comp
comp_1 comp_2 comp_3 anime_id name
0 36979.379560 19271.801823 -1.435758 8074.0 Highschool of the Dead
1 36864.979321 12890.819470 8.604092 8074.0 Highschool of the Dead
2 36501.255218 -8729.123562 -2.260925 8074.0 Highschool of the Dead
3 36504.981554 -7890.072244 -2.266146 8074.0 Highschool of the Dead
4 36575.509130 -2723.534636 -2.298376 8074.0 Highschool of the Dead
... ... ... ... ... ...
2330380 -36744.486077 -7441.741918 0.851111 17147.0 Gakkatsu! 2nd Season
2330381 -36660.120947 648.960605 -1.199318 25087.0 Ie Naki Ko Remi Specials (2001)
2330382 -36641.406527 1792.807522 -1.206458 29757.0 Mori no e
2330383 -36896.998157 -8205.349210 -2.143892 7616.0 Michi
2330384 -36959.758217 -7762.263236 -1.146544 28639.0 Futon

2330385 rows × 5 columns

Code 
comp_grp = comp.groupby('name').mean().reset_index()
comp_grp.name = comp_grp.name.str.replace('&quot;',"`")
comp_grp
name comp_1 comp_2 comp_3 anime_id
0 `0` -11120.206150 -5219.625973 -0.704695 20707.0
1 `Bungaku Shoujo` Kyou no Oyatsu: Hatsukoi 1635.899945 -518.535030 0.007314 7669.0
2 `Bungaku Shoujo` Memoire -1095.114790 -231.043774 0.065327 8481.0
3 `Bungaku Shoujo` Movie -1846.898975 40.053205 -0.135169 6408.0
4 `Eiji` -7032.249068 261.233554 1.754537 6076.0
... ... ... ... ... ...
8488 xxxHOLiC 1419.997992 93.118756 -0.038304 861.0
8489 xxxHOLiC Kei 2172.334257 179.512393 -0.058045 3091.0
8490 xxxHOLiC Movie: Manatsu no Yoru no Yume 3869.821060 436.500647 0.132941 793.0
8491 xxxHOLiC Rou -1700.534327 738.029495 -0.003196 6864.0
8492 xxxHOLiC Shunmuki 1465.196974 298.304491 0.074259 4918.0

8493 rows × 5 columns

Code 
import plotly.express as px
Code 
fig = px.scatter_3d(comp_grp, x='comp_1', y='comp_2', z='comp_3',color='name',size_max=7, opacity=0.7)
fig.update_layout(margin=dict(l=0, r=0, b=0, t=0))
fig.write_html('pca.html')

ELK Stack Recommendation System

Code 
anime['genre_list'] = anime.genre.str.split(',').apply(lambda x: [y.strip() for y in x])
Code 
anime[:3]
anime_id name genre type episodes rating members genre_list
0 32281 Kimi no Na wa. Drama, Romance, School, Supernatural Movie 1 9.37 200630 [Drama, Romance, School, Supernatural]
1 5114 Fullmetal Alchemist: Brotherhood Action, Adventure, Drama, Fantasy, Magic, Mili... TV 64 9.26 793665 [Action, Adventure, Drama, Fantasy, Magic, Mil...
2 28977 Gintama° Action, Comedy, Historical, Parody, Samurai, S... TV 51 9.25 114262 [Action, Comedy, Historical, Parody, Samurai, ...
Code 
rating[:3]
user_id anime_id rating
0 1 20 -1
1 1 24 -1
2 1 79 -1
Code 
try:
  import elasticsearch
except:
  !pip install elasticsearch
  import elasticsearch
Code 
from elasticsearch.helpers import bulk
Code 
import requests as req
Code 
es_host = ''
es_port = ''                                    
es_user = ''
es_pass = ''
Code 
print(req.get(f'{es_host}:{es_port}/_cat/indices',auth=(es_user, es_pass)).text)
green open apm-7.7.0-onboarding-2020.06.01 8AuzajpiSha3LKUIS1-0DA 1 1     1 0    14kb    7kb
green open apm-7.7.0-error-000001          g0mvvPSNRBqvTKT3XkxEtA 1 1     0 0    416b   208b
green open .apm-agent-configuration        NuHbtEHjReCuzcGzkUkTvQ 1 1     0 0    416b   208b
green open apm-7.7.0-transaction-000001    okf6EQuSRBWP_uPze2r0_g 1 1     0 0    416b   208b
green open .kibana_1                       9pXwv0ngRtaJYCuSSEmeRg 1 1   660 3   2.4mb  1.2mb
green open kibana_sample_data_flights      _JNZbj9KQnOL4cD8CIhfEw 1 1 13059 0  12.6mb  6.3mb
green open apm-7.7.0-span-000001           N8UCNrmaRjiaX8NnjHnlzw 1 1     0 0    416b   208b
green open .security-7                     cyFn8axBRwOjJDq5JHUhhQ 1 1    47 2 179.1kb 89.5kb
green open .apm-custom-link                NM_Cj4S4RCCzFNKFyyrwCQ 1 1     0 0    416b   208b
green open kibana_sample_data_ecommerce    ioSMO_biTUa2cusEJqScnA 1 1  4675 0   9.7mb  4.8mb
green open .kibana_task_manager_1          VHh1p97rQDCas32Xmi5LRA 1 1     5 1    83kb   20kb
green open apm-7.7.0-metric-000001         d9U7HlGqTYiXRSC_gZkrww 1 1     0 0    416b   208b
green open apm-7.7.0-profile-000001        TjVhTDXNQqaSW85kTRz3Ww 1 1     0 0    416b   208b
Code 
es = elasticsearch.Elasticsearch(
    [
        f'{es_host}:{es_port}'                          
    ], 
    http_auth=(es_user, es_pass),
  )
Code 
print(es.indices.get('*'))
Code 
def index_template(index,id,doc):
  return {
    '_index': index,
    '_id': id,
    '_source': doc
}
Code 
anime[:2]
anime_id name genre type episodes rating members genre_list
0 32281 Kimi no Na wa. Drama, Romance, School, Supernatural Movie 1 9.37 200630 [Drama, Romance, School, Supernatural]
1 5114 Fullmetal Alchemist: Brotherhood Action, Adventure, Drama, Fantasy, Magic, Mili... TV 64 9.26 793665 [Action, Adventure, Drama, Fantasy, Magic, Mil...
Code 
anime.anime_id.nunique() / len(anime)
1.0
Code 
anime_actions = [
  index_template('anime',x.anime_id,{
      'anime_id': x.anime_id ,
      'name': x.name,
      'genres': x.genre_list,
      'type': x.type if str(x.type) != 'nan' else 'Unknown',
      'episodes': int(x.episodes) if x.episodes not in ['Unknown'] else 0,
      'rating': x.rating if str(x.rating) != 'nan' else 0

  })  for x in anime.itertuples()              
]
anime_actions[:2]
[{'_id': 32281,
  '_index': 'anime',
  '_source': {'anime_id': 32281,
   'episodes': 1,
   'genres': ['Drama', 'Romance', 'School', 'Supernatural'],
   'name': 'Kimi no Na wa.',
   'rating': 9.37,
   'type': 'Movie'}},
 {'_id': 5114,
  '_index': 'anime',
  '_source': {'anime_id': 5114,
   'episodes': 64,
   'genres': ['Action',
    'Adventure',
    'Drama',
    'Fantasy',
    'Magic',
    'Military',
    'Shounen'],
   'name': 'Fullmetal Alchemist: Brotherhood',
   'rating': 9.26,
   'type': 'TV'}}]
Code 
%%time
bulk(es,anime_actions)
CPU times: user 306 ms, sys: 13.1 ms, total: 319 ms
Wall time: 8.6 s
(12294, [])
Code 
def positive(x):
  return 0 if x < 0 else x
Code 
frating = rating.groupby('user_id').agg({
    'anime_id': list,
    'rating': list
}).reset_index()
frating['mixed'] = frating[['anime_id','rating']].apply(lambda x: [ {'anime_id': i, 'rating': positive(r) if str(r) != 'nan' else 0} for i,r in zip(x.anime_id,x.rating)] ,axis=1)
frating
user_id anime_id rating mixed
0 1 [20, 24, 79, 226, 241, 355, 356, 442, 487, 846... [-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -... [{'anime_id': 20, 'rating': 0}, {'anime_id': 2...
1 2 [11771, 12189, 16417] [10, -1, -1] [{'anime_id': 11771, 'rating': 10}, {'anime_id...
2 3 [20, 154, 170, 199, 225, 341, 430, 527, 552, 8... [8, 6, 9, 10, 9, 6, 7, 7, 7, 10, 7, 7, 7, 8, 6... [{'anime_id': 20, 'rating': 8}, {'anime_id': 1...
3 4 [6, 72, 121, 150, 166, 205, 226, 857, 1292, 14... [-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -... [{'anime_id': 6, 'rating': 0}, {'anime_id': 72...
4 5 [6, 15, 17, 18, 20, 22, 24, 30, 45, 47, 57, 63... [8, 6, 6, 6, 6, 5, 1, 1, 7, 8, 7, 1, 6, 6, 6, ... [{'anime_id': 6, 'rating': 8}, {'anime_id': 15...
... ... ... ... ...
73510 73512 [60, 98, 101, 120, 199, 226, 339, 355, 552, 65... [10, 8, 10, 5, 10, 9, 10, -1, 7, 8, 8, 10, 8] [{'anime_id': 60, 'rating': 10}, {'anime_id': ...
73511 73513 [1, 5, 71, 101, 164, 180, 181, 195, 196, 199, ... [9, 8, 6, 8, 8, 5, 5, 7, 7, 9, 8, 8, 7, 10, 8,... [{'anime_id': 1, 'rating': 9}, {'anime_id': 5,...
73512 73514 [512] [10] [{'anime_id': 512, 'rating': 10}]
73513 73515 [1, 5, 6, 19, 27, 30, 33, 57, 67, 71, 72, 73, ... [10, 10, 10, 9, 9, 8, 10, 8, 8, 9, 10, 9, 10, ... [{'anime_id': 1, 'rating': 10}, {'anime_id': 5...
73514 73516 [790, 8074] [9, 9] [{'anime_id': 790, 'rating': 9}, {'anime_id': ...

73515 rows × 4 columns

Code 
user_actions = [
  index_template('anime-users',x.user_id,{
    'user_id': x.user_id,
    'rated': x.mixed
  })  for x in frating.itertuples()              
]
user_actions[:2]
[{'_id': 1,
  '_index': 'anime-users',
  '_source': {'rated': [{'anime_id': 20, 'rating': 0},
    {'anime_id': 24, 'rating': 0},
    {'anime_id': 79, 'rating': 0},
    {'anime_id': 226, 'rating': 0},
    {'anime_id': 241, 'rating': 0},
    {'anime_id': 355, 'rating': 0},
    {'anime_id': 356, 'rating': 0},
    {'anime_id': 442, 'rating': 0},
    {'anime_id': 487, 'rating': 0},
    {'anime_id': 846, 'rating': 0},
    {'anime_id': 936, 'rating': 0},
    {'anime_id': 1546, 'rating': 0},
    {'anime_id': 1692, 'rating': 0},
    {'anime_id': 1836, 'rating': 0},
    {'anime_id': 2001, 'rating': 0},
    {'anime_id': 2025, 'rating': 0},
    {'anime_id': 2144, 'rating': 0},
    {'anime_id': 2787, 'rating': 0},
    {'anime_id': 2993, 'rating': 0},
    {'anime_id': 3455, 'rating': 0},
    {'anime_id': 4063, 'rating': 0},
    {'anime_id': 4214, 'rating': 0},
    {'anime_id': 4224, 'rating': 0},
    {'anime_id': 4581, 'rating': 0},
    {'anime_id': 4744, 'rating': 0},
    {'anime_id': 4898, 'rating': 0},
    {'anime_id': 4999, 'rating': 0},
    {'anime_id': 5034, 'rating': 0},
    {'anime_id': 5277, 'rating': 0},
    {'anime_id': 5667, 'rating': 0},
    {'anime_id': 5781, 'rating': 0},
    {'anime_id': 5958, 'rating': 0},
    {'anime_id': 6163, 'rating': 0},
    {'anime_id': 6205, 'rating': 0},
    {'anime_id': 6324, 'rating': 0},
    {'anime_id': 6500, 'rating': 0},
    {'anime_id': 6547, 'rating': 0},
    {'anime_id': 6682, 'rating': 0},
    {'anime_id': 6707, 'rating': 0},
    {'anime_id': 6747, 'rating': 0},
    {'anime_id': 6773, 'rating': 0},
    {'anime_id': 6793, 'rating': 0},
    {'anime_id': 7088, 'rating': 0},
    {'anime_id': 7148, 'rating': 0},
    {'anime_id': 7593, 'rating': 0},
    {'anime_id': 7739, 'rating': 0},
    {'anime_id': 7858, 'rating': 0},
    {'anime_id': 8074, 'rating': 10},
    {'anime_id': 8407, 'rating': 0},
    {'anime_id': 8424, 'rating': 0},
    {'anime_id': 8525, 'rating': 0},
    {'anime_id': 8630, 'rating': 0},
    {'anime_id': 8841, 'rating': 0},
    {'anime_id': 9041, 'rating': 0},
    {'anime_id': 9062, 'rating': 0},
    {'anime_id': 9136, 'rating': 0},
    {'anime_id': 9181, 'rating': 0},
    {'anime_id': 9330, 'rating': 0},
    {'anime_id': 9367, 'rating': 0},
    {'anime_id': 9515, 'rating': 0},
    {'anime_id': 9581, 'rating': 0},
    {'anime_id': 9675, 'rating': 0},
    {'anime_id': 9750, 'rating': 0},
    {'anime_id': 9790, 'rating': 0},
    {'anime_id': 9919, 'rating': 0},
    {'anime_id': 10067, 'rating': 0},
    {'anime_id': 10073, 'rating': 0},
    {'anime_id': 10076, 'rating': 0},
    {'anime_id': 10079, 'rating': 0},
    {'anime_id': 10080, 'rating': 0},
    {'anime_id': 10209, 'rating': 0},
    {'anime_id': 10578, 'rating': 0},
    {'anime_id': 10604, 'rating': 0},
    {'anime_id': 10719, 'rating': 0},
    {'anime_id': 10790, 'rating': 0},
    {'anime_id': 10793, 'rating': 0},
    {'anime_id': 10794, 'rating': 0},
    {'anime_id': 10805, 'rating': 0},
    {'anime_id': 10897, 'rating': 0},
    {'anime_id': 11161, 'rating': 0},
    {'anime_id': 11266, 'rating': 0},
    {'anime_id': 11617, 'rating': 10},
    {'anime_id': 11737, 'rating': 0},
    {'anime_id': 11757, 'rating': 10},
    {'anime_id': 11759, 'rating': 0},
    {'anime_id': 11771, 'rating': 0},
    {'anime_id': 12293, 'rating': 0},
    {'anime_id': 12549, 'rating': 0},
    {'anime_id': 12729, 'rating': 0},
    {'anime_id': 13357, 'rating': 0},
    {'anime_id': 13367, 'rating': 0},
    {'anime_id': 13411, 'rating': 0},
    {'anime_id': 13561, 'rating': 0},
    {'anime_id': 13663, 'rating': 0},
    {'anime_id': 13759, 'rating': 0},
    {'anime_id': 14749, 'rating': 0},
    {'anime_id': 14813, 'rating': 0},
    {'anime_id': 14833, 'rating': 0},
    {'anime_id': 14967, 'rating': 0},
    {'anime_id': 15117, 'rating': 0},
    {'anime_id': 15437, 'rating': 0},
    {'anime_id': 15451, 'rating': 10},
    {'anime_id': 15583, 'rating': 0},
    {'anime_id': 15609, 'rating': 0},
    {'anime_id': 16011, 'rating': 0},
    {'anime_id': 16498, 'rating': 0},
    {'anime_id': 16706, 'rating': 0},
    {'anime_id': 17265, 'rating': 0},
    {'anime_id': 17729, 'rating': 0},
    {'anime_id': 18247, 'rating': 0},
    {'anime_id': 18277, 'rating': 0},
    {'anime_id': 18753, 'rating': 0},
    {'anime_id': 18897, 'rating': 0},
    {'anime_id': 19163, 'rating': 0},
    {'anime_id': 19221, 'rating': 0},
    {'anime_id': 19285, 'rating': 0},
    {'anime_id': 19429, 'rating': 0},
    {'anime_id': 19815, 'rating': 0},
    {'anime_id': 20045, 'rating': 0},
    {'anime_id': 20785, 'rating': 0},
    {'anime_id': 20787, 'rating': 0},
    {'anime_id': 21033, 'rating': 0},
    {'anime_id': 21881, 'rating': 0},
    {'anime_id': 22147, 'rating': 0},
    {'anime_id': 22199, 'rating': 0},
    {'anime_id': 22319, 'rating': 0},
    {'anime_id': 22535, 'rating': 0},
    {'anime_id': 22547, 'rating': 0},
    {'anime_id': 22663, 'rating': 0},
    {'anime_id': 22877, 'rating': 0},
    {'anime_id': 23233, 'rating': 0},
    {'anime_id': 23321, 'rating': 0},
    {'anime_id': 23847, 'rating': 0},
    {'anime_id': 24133, 'rating': 0},
    {'anime_id': 24455, 'rating': 0},
    {'anime_id': 24873, 'rating': 0},
    {'anime_id': 25099, 'rating': 0},
    {'anime_id': 25157, 'rating': 0},
    {'anime_id': 25159, 'rating': 0},
    {'anime_id': 25283, 'rating': 0},
    {'anime_id': 25397, 'rating': 0},
    {'anime_id': 26243, 'rating': 0},
    {'anime_id': 27775, 'rating': 0},
    {'anime_id': 27899, 'rating': 0},
    {'anime_id': 28121, 'rating': 0},
    {'anime_id': 28677, 'rating': 0},
    {'anime_id': 29093, 'rating': 0},
    {'anime_id': 29095, 'rating': 0},
    {'anime_id': 30015, 'rating': 0},
    {'anime_id': 30296, 'rating': 0},
    {'anime_id': 30544, 'rating': 0},
    {'anime_id': 31338, 'rating': 0},
    {'anime_id': 31845, 'rating': 0}],
   'user_id': 1}},
 {'_id': 2,
  '_index': 'anime-users',
  '_source': {'rated': [{'anime_id': 11771, 'rating': 10},
    {'anime_id': 12189, 'rating': 0},
    {'anime_id': 16417, 'rating': 0}],
   'user_id': 2}}]
Code 
%%time
bulk(es,user_actions)
CPU times: user 8.01 s, sys: 166 ms, total: 8.18 s
Wall time: 1min 41s
(73515, [])