simple twitter approach

This commit is contained in:
Jonas Weinz 2018-06-05 15:51:58 +02:00
parent b6ed452fc6
commit 26c3c2abf1
2 changed files with 18486 additions and 8 deletions

View File

@ -2,7 +2,7 @@
"cells": [ "cells": [
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 13, "execution_count": 22,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -11,7 +11,13 @@
"import ipywidgets as widgets\n", "import ipywidgets as widgets\n",
"import os\n", "import os\n",
"import glob\n", "import glob\n",
"import json" "import json\n",
"import numpy as np\n",
"import itertools\n",
"import sklearn.utils as sku\n",
"from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer, HashingVectorizer\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.preprocessing import MultiLabelBinarizer"
] ]
}, },
{ {
@ -31,7 +37,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 9, "execution_count": 3,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -47,7 +53,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 10, "execution_count": 4,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -70,7 +76,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 16, "execution_count": 5,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
@ -912,7 +918,7 @@
"[176332 rows x 7 columns]" "[176332 rows x 7 columns]"
] ]
}, },
"execution_count": 16, "execution_count": 5,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -931,7 +937,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 20, "execution_count": 6,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -950,10 +956,11 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 21, "execution_count": 7,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"# defining blacklist for modifier emojis:\n",
"emoji_blacklist = set([\n", "emoji_blacklist = set([\n",
" chr(0x1F3FB),\n", " chr(0x1F3FB),\n",
" chr(0x1F3FC),\n", " chr(0x1F3FC),\n",
@ -964,6 +971,813 @@
" chr(0x2640)\n", " chr(0x2640)\n",
"])" "])"
] ]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [],
"source": [
"# filtering them and the EMOJI keyword out:\n",
"plain_text = plain_text.str.replace(\"<EMOJI>\",\"\").str.replace(\"[\" + \"\".join(list(emoji_blacklist)) + \"]\",\"\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# making a set of the emoji list:\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"* trying a multi label vectorizer"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [],
"source": [
"mlb = MultiLabelBinarizer()\n",
"\n",
"labels = mlb.fit_transform(emojis)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"* generating train and test set:"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [],
"source": [
"X1, Xt1, y1, yt1 = train_test_split(plain_text, labels, test_size=0.1, random_state=4222)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [],
"source": [
"vectorizer = TfidfVectorizer(stop_words='english')\n",
"vec_train = vectorizer.fit_transform(X1)\n",
"vec_test = vectorizer.transform(Xt1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"* train. this can take a very long time..."
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.neural_network import MLPClassifier as MLP\n",
"from sklearn.multiclass import OneVsRestClassifier as OVRC\n",
"from sklearn.tree import DecisionTreeClassifier as DTC"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {},
"outputs": [],
"source": [
"def train():\n",
" clf = OVRC(DTC())\n",
"\n",
" clf.fit(vec_train[:10000], y1[:10000])\n",
" \n",
" return clf"
]
},
{
"cell_type": "code",
"execution_count": 50,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 889 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 897 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 898 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 901 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 902 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 903 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 907 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 908 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 909 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 910 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 911 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 913 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 914 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 915 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 918 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 919 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 924 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 925 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 926 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 927 is present in all training examples.\n",
" str(classes[c]))\n",
"/home/jonas/.local/lib/python3.6/site-packages/sklearn/multiclass.py:76: UserWarning: Label not 928 is present in all training examples.\n",
" str(classes[c]))\n"
]
}
],
"source": [
"clf = train()"
]
},
{
"cell_type": "code",
"execution_count": 51,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>predict</th>\n",
" <th>teacher</th>\n",
" <th>text</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>57624</th>\n",
" <td>(💥, 😡)</td>\n",
" <td>(💥, 😡)</td>\n",
" <td>@RealRoseTaylor1: IMO A terrorist doesnt hav...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>58180</th>\n",
" <td>()</td>\n",
" <td>(😒,)</td>\n",
" <td>@mog7546: NOT BREAKING NEWS - #KELLY's STUPID...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>122719</th>\n",
" <td>(😂, 🤦)</td>\n",
" <td>(😂, 🤦)</td>\n",
" <td>@SpecialMonroe: Im more of like a “ yes nigg...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>132425</th>\n",
" <td>()</td>\n",
" <td>(🤷,)</td>\n",
" <td>@xo_sweetttnesss Basically </td>\n",
" </tr>\n",
" <tr>\n",
" <th>95629</th>\n",
" <td>(😂,)</td>\n",
" <td>(😂,)</td>\n",
" <td>@JVino_: Never should of gave niggas technolo...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>144805</th>\n",
" <td>(😂,)</td>\n",
" <td>(🤷,)</td>\n",
" <td>@_Erickaaaaaaa: People mad iffy Im not aroun...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>50997</th>\n",
" <td>(💕, 🛫)</td>\n",
" <td>(💕, 🛫)</td>\n",
" <td>@MURASAKI_9394: &amp;lt;preview&amp;gt; 171101 ICN@mt...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>52294</th>\n",
" <td>()</td>\n",
" <td>(😍, 😣)</td>\n",
" <td>@JaDinePhils: Rogue x Gambit x Deadpool No cu...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>57600</th>\n",
" <td>(⚾, 😬)</td>\n",
" <td>(💙,)</td>\n",
" <td>@CailenHanzlik: So hyped watching my former t...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12709</th>\n",
" <td>(😂,)</td>\n",
" <td>(💀,)</td>\n",
" <td>@C_eazyy @RuthlessLuke_ LMAO DED let a nigga s...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>159921</th>\n",
" <td>(😍,)</td>\n",
" <td>(👎,)</td>\n",
" <td>@ShawnHayes_Real @Mr_Ghostly @CR_Shelton @Rojo...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1297</th>\n",
" <td>(💀, 😂)</td>\n",
" <td>(💀, 😂)</td>\n",
" <td>@Ieansquad: he really solved the equation I c...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>124445</th>\n",
" <td>()</td>\n",
" <td>(😩,)</td>\n",
" <td>It is really breaking my heart. Imma stick by ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>150159</th>\n",
" <td>()</td>\n",
" <td>(🐰,)</td>\n",
" <td>@jikookarchive: : aw sexy~: (jimin your cleav...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>170959</th>\n",
" <td>(✅, 🔥)</td>\n",
" <td>(✅, 🍋, 🔥)</td>\n",
" <td>@Harrys1DEmpire: 1: Like this2: Follow all th...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>46992</th>\n",
" <td>(🎬, 💋, 🔗)</td>\n",
" <td>(🎬, 💋, 🔗)</td>\n",
" <td>@ONGSEONGWU_TH: [] #WANNAONE X Innisfree | On...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>130772</th>\n",
" <td>(🎃,)</td>\n",
" <td>(🐺,)</td>\n",
" <td>@lbtreiman: Well i certainly didn't get 'em a...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>63733</th>\n",
" <td>()</td>\n",
" <td>(🎃,)</td>\n",
" <td>@gamngbizzle: Follow everyone who retweets t...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>22026</th>\n",
" <td>()</td>\n",
" <td>(💕, 📷)</td>\n",
" <td>@fshion_me: Love it【@perlinek】Coupon code: zo...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>109588</th>\n",
" <td>(😍, 😳)</td>\n",
" <td>(😍,)</td>\n",
" <td>@SoCuteBabies: Those eyes</td>\n",
" </tr>\n",
" <tr>\n",
" <th>167358</th>\n",
" <td>(🎃,)</td>\n",
" <td>(👻, 😈)</td>\n",
" <td>@Lepetit626: Happy Halloween from the cutest ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>132276</th>\n",
" <td>()</td>\n",
" <td>(💀,)</td>\n",
" <td>@marceoooo shuttup</td>\n",
" </tr>\n",
" <tr>\n",
" <th>150897</th>\n",
" <td>(😱, 🙈, 🙉)</td>\n",
" <td>(🔥, 😈)</td>\n",
" <td>@instaWANKtube: Drilling babe@mistersex17 @hq...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>15211</th>\n",
" <td>(🎃,)</td>\n",
" <td>(🎃,)</td>\n",
" <td>@G_Eazy: Hopped out of the millennium falcon ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>58157</th>\n",
" <td>(⌛, 💊)</td>\n",
" <td>(⌛, 💊)</td>\n",
" <td>@NBCNewsNight: Stranger Things Neuro Upside D...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>87036</th>\n",
" <td>(🔥,)</td>\n",
" <td>(😂,)</td>\n",
" <td>I ran a lot harder back then, as well.I join v...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>68909</th>\n",
" <td>()</td>\n",
" <td>(🚗,)</td>\n",
" <td>@Holy_Boost: SUMMER BOOST case GA! -Like+ ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>70063</th>\n",
" <td>()</td>\n",
" <td>(😭,)</td>\n",
" <td>I love you so much meliflua ❤️</td>\n",
" </tr>\n",
" <tr>\n",
" <th>53508</th>\n",
" <td>()</td>\n",
" <td>(💥,)</td>\n",
" <td>@SMMmarkeing786: Get FREE CBD OIL Trial Bott...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>91590</th>\n",
" <td>(👀,)</td>\n",
" <td>(👀,)</td>\n",
" <td>@dezidoesit: Shea Moisture did not come to pl...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>...</th>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>82637</th>\n",
" <td>()</td>\n",
" <td>()</td>\n",
" <td>@KissaSins: Follow me on Instagram!!</td>\n",
" </tr>\n",
" <tr>\n",
" <th>151805</th>\n",
" <td>(👀,)</td>\n",
" <td>(👀,)</td>\n",
" <td>@dezidoesit: Shea Moisture did not come to pl...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>25914</th>\n",
" <td>()</td>\n",
" <td>(🔘,)</td>\n",
" <td>@Heartless600904: He knows how badly he hurt ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>91602</th>\n",
" <td>()</td>\n",
" <td>(😡,)</td>\n",
" <td>@1776Stonewall @Frederick99m That's when I kne...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>16911</th>\n",
" <td>()</td>\n",
" <td>(😩,)</td>\n",
" <td>Not even out of bed yet and Im ready to get o...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>94798</th>\n",
" <td>()</td>\n",
" <td>(🙃,)</td>\n",
" <td>nobody gets how lonely and hungry i get</td>\n",
" </tr>\n",
" <tr>\n",
" <th>112317</th>\n",
" <td>(💪,)</td>\n",
" <td>(🤗,)</td>\n",
" <td>@BIGAC_: new month, new money, new opportunit...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>72241</th>\n",
" <td>()</td>\n",
" <td>(😎,)</td>\n",
" <td>@ZeeTamil: Happy birthday @iamsrk #ZeeTamil #...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>160551</th>\n",
" <td>(✨, 💪)</td>\n",
" <td>(😭,)</td>\n",
" <td>@WORLDSTARNOW: Drake looks like he entering h...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>151889</th>\n",
" <td>(👅, 😻)</td>\n",
" <td>(😍,)</td>\n",
" <td>@GenderReveaIs: The moment of truth, hoop sty...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>64998</th>\n",
" <td>(🐶, 💓)</td>\n",
" <td>(🐶, 💓)</td>\n",
" <td>@LA_Loving: Babys first Halloween</td>\n",
" </tr>\n",
" <tr>\n",
" <th>138015</th>\n",
" <td>()</td>\n",
" <td>(😂,)</td>\n",
" <td>@MikeChosen1: When you got a girl who doesn't...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>84447</th>\n",
" <td>()</td>\n",
" <td>(💕,)</td>\n",
" <td>@LiamPayne @pinkskiesljp @BestSimy @LiamPayne ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>125899</th>\n",
" <td>()</td>\n",
" <td>(🙏,)</td>\n",
" <td>@Oh_Ashlaay We just have to pray to the baseba...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>47898</th>\n",
" <td>(😂,)</td>\n",
" <td>(😂,)</td>\n",
" <td>@JVino_: Never should of gave niggas technolo...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7597</th>\n",
" <td>()</td>\n",
" <td>(😂,)</td>\n",
" <td>@junkiejunkrat Were doing goooooood. But fuck...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>173305</th>\n",
" <td>()</td>\n",
" <td>(🌈, 💛)</td>\n",
" <td>Ohh nice.</td>\n",
" </tr>\n",
" <tr>\n",
" <th>133827</th>\n",
" <td>(💙,)</td>\n",
" <td>(💙,)</td>\n",
" <td>@JenBluesnake: This is lovely. Unsworth's app...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>157487</th>\n",
" <td>()</td>\n",
" <td>(💻,)</td>\n",
" <td>@yesyours: : Vehicle Magnet #VehicleMagnet 🖨</td>\n",
" </tr>\n",
" <tr>\n",
" <th>142977</th>\n",
" <td>(🌸, 💓)</td>\n",
" <td>(🌸, 💓)</td>\n",
" <td>@namkook_daily: they are so cosy and soft ☁</td>\n",
" </tr>\n",
" <tr>\n",
" <th>47506</th>\n",
" <td>(😭,)</td>\n",
" <td>(💙, 🔥)</td>\n",
" <td>@Juli_SJ13: ••••••2017 SAVED ••••••#BlackSuit</td>\n",
" </tr>\n",
" <tr>\n",
" <th>164518</th>\n",
" <td>()</td>\n",
" <td>(😜,)</td>\n",
" <td>@marcogumabao: I hope you like your eggs SUNN...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>96549</th>\n",
" <td>(🐶, 💓)</td>\n",
" <td>(🐶, 💓)</td>\n",
" <td>@LA_Loving: Babys first Halloween</td>\n",
" </tr>\n",
" <tr>\n",
" <th>69134</th>\n",
" <td>()</td>\n",
" <td>(😞, 😡)</td>\n",
" <td>Reading this @AshcroftBen with many mixed emot...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>116753</th>\n",
" <td>(💪,)</td>\n",
" <td>(🍋,)</td>\n",
" <td>@Caradelevingne: ❤️ NO_ONE EVER REALLY DIES ❤...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>125133</th>\n",
" <td>(😑,)</td>\n",
" <td>(🖤, 😉)</td>\n",
" <td>@DannyTu3250187: @Kriszti7504 @filipina_krueg...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5081</th>\n",
" <td>(😔,)</td>\n",
" <td>(😘,)</td>\n",
" <td>@brysoxox: rt or gay</td>\n",
" </tr>\n",
" <tr>\n",
" <th>62088</th>\n",
" <td>()</td>\n",
" <td>(💔,)</td>\n",
" <td>Why do say that hondas are for fboys??</td>\n",
" </tr>\n",
" <tr>\n",
" <th>32769</th>\n",
" <td>(😂, 😘, 🙏)</td>\n",
" <td>(😁,)</td>\n",
" <td>@menggalurks: Live na ba ang EB today? #ALDU...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>92385</th>\n",
" <td>(😍,)</td>\n",
" <td>(💔,)</td>\n",
" <td>OML, WHAT HAPPENED?!!</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>17634 rows × 3 columns</p>\n",
"</div>"
],
"text/plain": [
" predict teacher \\\n",
"57624 (💥, 😡) (💥, 😡) \n",
"58180 () (😒,) \n",
"122719 (😂, 🤦) (😂, 🤦) \n",
"132425 () (🤷,) \n",
"95629 (😂,) (😂,) \n",
"144805 (😂,) (🤷,) \n",
"50997 (💕, 🛫) (💕, 🛫) \n",
"52294 () (😍, 😣) \n",
"57600 (⚾, 😬) (💙,) \n",
"12709 (😂,) (💀,) \n",
"159921 (😍,) (👎,) \n",
"1297 (💀, 😂) (💀, 😂) \n",
"124445 () (😩,) \n",
"150159 () (🐰,) \n",
"170959 (✅, 🔥) (✅, 🍋, 🔥) \n",
"46992 (🎬, 💋, 🔗) (🎬, 💋, 🔗) \n",
"130772 (🎃,) (🐺,) \n",
"63733 () (🎃,) \n",
"22026 () (💕, 📷) \n",
"109588 (😍, 😳) (😍,) \n",
"167358 (🎃,) (👻, 😈) \n",
"132276 () (💀,) \n",
"150897 (😱, 🙈, 🙉) (🔥, 😈) \n",
"15211 (🎃,) (🎃,) \n",
"58157 (⌛, 💊) (⌛, 💊) \n",
"87036 (🔥,) (😂,) \n",
"68909 () (🚗,) \n",
"70063 () (😭,) \n",
"53508 () (💥,) \n",
"91590 (👀,) (👀,) \n",
"... ... ... \n",
"82637 () () \n",
"151805 (👀,) (👀,) \n",
"25914 () (🔘,) \n",
"91602 () (😡,) \n",
"16911 () (😩,) \n",
"94798 () (🙃,) \n",
"112317 (💪,) (🤗,) \n",
"72241 () (😎,) \n",
"160551 (✨, 💪) (😭,) \n",
"151889 (👅, 😻) (😍,) \n",
"64998 (🐶, 💓) (🐶, 💓) \n",
"138015 () (😂,) \n",
"84447 () (💕,) \n",
"125899 () (🙏,) \n",
"47898 (😂,) (😂,) \n",
"7597 () (😂,) \n",
"173305 () (🌈, 💛) \n",
"133827 (💙,) (💙,) \n",
"157487 () (💻,) \n",
"142977 (🌸, 💓) (🌸, 💓) \n",
"47506 (😭,) (💙, 🔥) \n",
"164518 () (😜,) \n",
"96549 (🐶, 💓) (🐶, 💓) \n",
"69134 () (😞, 😡) \n",
"116753 (💪,) (🍋,) \n",
"125133 (😑,) (🖤, 😉) \n",
"5081 (😔,) (😘,) \n",
"62088 () (💔,) \n",
"32769 (😂, 😘, 🙏) (😁,) \n",
"92385 (😍,) (💔,) \n",
"\n",
" text \n",
"57624 @RealRoseTaylor1: IMO A terrorist doesnt hav... \n",
"58180 @mog7546: NOT BREAKING NEWS - #KELLY's STUPID... \n",
"122719 @SpecialMonroe: Im more of like a “ yes nigg... \n",
"132425 @xo_sweetttnesss Basically \n",
"95629 @JVino_: Never should of gave niggas technolo... \n",
"144805 @_Erickaaaaaaa: People mad iffy Im not aroun... \n",
"50997 @MURASAKI_9394: &lt;preview&gt; 171101 ICN@mt... \n",
"52294 @JaDinePhils: Rogue x Gambit x Deadpool No cu... \n",
"57600 @CailenHanzlik: So hyped watching my former t... \n",
"12709 @C_eazyy @RuthlessLuke_ LMAO DED let a nigga s... \n",
"159921 @ShawnHayes_Real @Mr_Ghostly @CR_Shelton @Rojo... \n",
"1297 @Ieansquad: he really solved the equation I c... \n",
"124445 It is really breaking my heart. Imma stick by ... \n",
"150159 @jikookarchive: : aw sexy~: (jimin your cleav... \n",
"170959 @Harrys1DEmpire: 1: Like this2: Follow all th... \n",
"46992 @ONGSEONGWU_TH: [] #WANNAONE X Innisfree | On... \n",
"130772 @lbtreiman: Well i certainly didn't get 'em a... \n",
"63733 @gamngbizzle: Follow everyone who retweets t... \n",
"22026 @fshion_me: Love it【@perlinek】Coupon code: zo... \n",
"109588 @SoCuteBabies: Those eyes \n",
"167358 @Lepetit626: Happy Halloween from the cutest ... \n",
"132276 @marceoooo shuttup \n",
"150897 @instaWANKtube: Drilling babe@mistersex17 @hq... \n",
"15211 @G_Eazy: Hopped out of the millennium falcon ... \n",
"58157 @NBCNewsNight: Stranger Things Neuro Upside D... \n",
"87036 I ran a lot harder back then, as well.I join v... \n",
"68909 @Holy_Boost: SUMMER BOOST case GA! -Like+ ... \n",
"70063 I love you so much meliflua ❤️ \n",
"53508 @SMMmarkeing786: Get FREE CBD OIL Trial Bott... \n",
"91590 @dezidoesit: Shea Moisture did not come to pl... \n",
"... ... \n",
"82637 @KissaSins: Follow me on Instagram!! \n",
"151805 @dezidoesit: Shea Moisture did not come to pl... \n",
"25914 @Heartless600904: He knows how badly he hurt ... \n",
"91602 @1776Stonewall @Frederick99m That's when I kne... \n",
"16911 Not even out of bed yet and Im ready to get o... \n",
"94798 nobody gets how lonely and hungry i get \n",
"112317 @BIGAC_: new month, new money, new opportunit... \n",
"72241 @ZeeTamil: Happy birthday @iamsrk #ZeeTamil #... \n",
"160551 @WORLDSTARNOW: Drake looks like he entering h... \n",
"151889 @GenderReveaIs: The moment of truth, hoop sty... \n",
"64998 @LA_Loving: Babys first Halloween \n",
"138015 @MikeChosen1: When you got a girl who doesn't... \n",
"84447 @LiamPayne @pinkskiesljp @BestSimy @LiamPayne ... \n",
"125899 @Oh_Ashlaay We just have to pray to the baseba... \n",
"47898 @JVino_: Never should of gave niggas technolo... \n",
"7597 @junkiejunkrat Were doing goooooood. But fuck... \n",
"173305 Ohh nice. \n",
"133827 @JenBluesnake: This is lovely. Unsworth's app... \n",
"157487 @yesyours: : Vehicle Magnet #VehicleMagnet 🖨 \n",
"142977 @namkook_daily: they are so cosy and soft ☁ \n",
"47506 @Juli_SJ13: ••••••2017 SAVED ••••••#BlackSuit \n",
"164518 @marcogumabao: I hope you like your eggs SUNN... \n",
"96549 @LA_Loving: Babys first Halloween \n",
"69134 Reading this @AshcroftBen with many mixed emot... \n",
"116753 @Caradelevingne: ❤️ NO_ONE EVER REALLY DIES ❤... \n",
"125133 @DannyTu3250187: @Kriszti7504 @filipina_krueg... \n",
"5081 @brysoxox: rt or gay \n",
"62088 Why do say that hondas are for fboys?? \n",
"32769 @menggalurks: Live na ba ang EB today? #ALDU... \n",
"92385 OML, WHAT HAPPENED?!! \n",
"\n",
"[17634 rows x 3 columns]"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"pred = clf.predict(vectorizer.transform(Xt1))\n",
"\n",
"# build a dataframe to visualize test results:\n",
"testlist = pd.DataFrame({'text': Xt1, 'teacher': mlb.inverse_transform(yt1), 'predict': mlb.inverse_transform(pred)})\n",
"\n",
"display(testlist)"
]
},
{
"cell_type": "code",
"execution_count": 52,
"metadata": {},
"outputs": [],
"source": [
"testlist.to_csv('test.csv')"
]
},
{
"cell_type": "code",
"execution_count": 58,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "1e6d156f9beb42ada3790d977beb2b7c",
"version_major": 2,
"version_minor": 0
},
"text/html": [
"<p>Failed to display Jupyter Widget of type <code>Text</code>.</p>\n",
"<p>\n",
" If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
" that the widgets JavaScript is still loading. If this message persists, it\n",
" likely means that the widgets JavaScript library is either not installed or\n",
" not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
" Widgets Documentation</a> for setup instructions.\n",
"</p>\n",
"<p>\n",
" If you're reading this message in another frontend (for example, a static\n",
" rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
" it may mean that your frontend doesn't currently support widgets.\n",
"</p>\n"
],
"text/plain": [
"Text(value='')"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "b76a73c0fa7946f28fe91e18240484a4",
"version_major": 2,
"version_minor": 0
},
"text/html": [
"<p>Failed to display Jupyter Widget of type <code>VBox</code>.</p>\n",
"<p>\n",
" If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
" that the widgets JavaScript is still loading. If this message persists, it\n",
" likely means that the widgets JavaScript library is either not installed or\n",
" not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
" Widgets Documentation</a> for setup instructions.\n",
"</p>\n",
"<p>\n",
" If you're reading this message in another frontend (for example, a static\n",
" rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
" it may mean that your frontend doesn't currently support widgets.\n",
"</p>\n"
],
"text/plain": [
"VBox(children=(Button(description='get smiley', icon='check', style=ButtonStyle(), tooltip='Click me'), Output()))"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"out = widgets.Output()\n",
"\n",
"t = widgets.Text()\n",
"b = widgets.Button(\n",
" description='get smiley',\n",
" disabled=False,\n",
" button_style='', # 'success', 'info', 'warning', 'danger' or ''\n",
" tooltip='Click me',\n",
" icon='check'\n",
")\n",
"\n",
"\n",
"\n",
"def handle_submit(sender):\n",
" with out:\n",
" clear_output()\n",
" with out:\n",
" display(Markdown(\"# \" + str(mlb.inverse_transform(clf.predict(vectorizer.transform([t.value])))[0])))\n",
"\n",
"b.on_click(handle_submit)\n",
" \n",
"display(t)\n",
"display(widgets.VBox([b, out])) "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
} }
], ],
"metadata": { "metadata": {

File diff suppressed because it is too large Load Diff