The Emojis of Kathmandu

50K tweets between 2012 to 2021 (containing emojis, and geocoded to Kathmandu) show that we really love 😂

Recently though, we have been 😂 less and 🙏 more, 😁 less and ❤️ more. Just goes to show Kathmandu is healing from Covid.

Github repo: https://github.com/ayushsubedi/emojan

Create the dataset using twint

import twint
import pandas as pd
import nest_asyncio
nest_asyncio.apply()

emoji_list = ['😁',
 '😒',
 '😊',
 '😈',
 '🇳🇵',
 '😌',
 '☕',
 '👶',
 '👍',
 '😷',
 '👌',
 '🌞',
 '😑',
 '😉',
 '😍',
 '☺️',
 '😴',
 '😱',
 '🙏',
 '😘',
 '🙌',
 '😔',
 '😋',
 '😂',
 '😩',
 '💕',
 '😎',
 '😭',
 '😳',
 '😇',
 '😏',
 '😜',
 '☕️',
 '✌️',
 '🙈',
 '❤️',
 '😄',
 '💞']

for emoji in emoji_list:
    c = twint.Config()
    c.Search = emoji
    c.Pandas = True
    c.Store_csv = True
    c.Output = emoji
    c.Hide_output= True
    c.Near= "kathmandu"
    c.Since="2010-01-01"
    c.Until = "2021-05-31"
    twint.run.Search(c)

Perform analysis

import pandas as pd

df = pd.read_csv("../datasets/merged.csv", lineterminator='\n', parse_dates=['date'])

list(df)

['id', 'emoji', 'date', 'username', 'tweet', 'likes_count', 'place']

df.username.value_counts().head(20)

ms_madhur          1433
sristee44          1121
beingsamikshya      871
anuskashresthax     618
nepalplanettrek     594
scousergirl         572
dreamingdr          569
nepaligentleman     476
itsme_shivangi      433
iamnabinraj75       410
sim_shrestha        402
milan_pu1           353
raunakbasnet1       332
thulokanxo          325
chetan_karki        302
saampokhrel         301
mongolianheartk     300
tenzintsetenbhu     290
rana1997rohit       275
fatyangri           256
Name: username, dtype: int64

df.emoji.value_counts().head(20)

😂     7576
😊     4016
😍     3781
😁     3482
🇳🇵    2785
😎     2270
❤️    2233
🙏     2122
😉     2000
😜     1463
😘     1432
😋     1241
💕     1173
😄     1095
👍     1001
👌      730
😭      637
😏      554
✌️     505
😒      483
Name: emoji, dtype: int64

df.date.min()

Timestamp('2012-01-28 00:00:00')

df.set_index('date').resample('M')['id'].count()

date
2012-01-31      2
2012-02-29     16
2012-03-31     13
2012-04-30      3
2012-05-31     11
             ... 
2021-01-31    212
2021-02-28    134
2021-03-31    183
2021-04-30    160
2021-05-31    119
Freq: M, Name: id, Length: 113, dtype: int64

df.set_index('date').resample('M')['id'].count().plot()

<AxesSubplot:xlabel='date'>

df[df.emoji=="❤️"].set_index('date').resample('M')['id'].count().plot()

<AxesSubplot:xlabel='date'>

df[df.emoji=="😷"].set_index('date').resample('M')['id'].count().plot()

<AxesSubplot:xlabel='date'>

df[df.emoji=="🇳🇵"].set_index('date').resample('M')['id'].count().plot()

<AxesSubplot:xlabel='date'>

df[df.emoji=="😂"].set_index('date').resample('M')['id'].count().plot()

<AxesSubplot:xlabel='date'>

location_tweet = df.dropna(subset=['place'])

import json
import math

def getlatlon(row):
    place = row.place
    place = place.replace("\'", "\"")
    place = json.loads(place)
    lat = place['coordinates'][0]
    lon = place['coordinates'][1]
    return pd.Series([lat, lon],index=['lat','lon'])

df = df.join(location_tweet.apply(getlatlon, axis=1, result_type="expand"))

df.to_csv('location.csv', index=False)

df

df.emoji.value_counts()

😂     7576
😊     4016
😍     3781
😁     3482
🇳🇵    2785
😎     2270
❤️    2233
🙏     2122
😉     2000
😜     1463
😘     1432
😋     1241
💕     1173
😄     1095
👍     1001
👌      730
😭      637
😏      554
✌️     505
😒      483
💞      461
😇      432
🙌      423
☺️     375
😔      373
😌      332
😱      314
😑      296
😈      292
🙈      291
☕      277
🌞      213
😳      185
😩      181
😴      163
😷      135
👶       45
Name: emoji, dtype: int64