From ef5422bb0c8ac594d02299f0c80eef4977903c80 Mon Sep 17 00:00:00 2001 From: Jonas Weinz Date: Sat, 19 May 2018 20:43:06 +0200 Subject: [PATCH] more emoji stuff --- Project/Tools/emoji tester.ipynb | 2590 +++--------------------------- Project/Tools/whatsapp2csv.sh | 2 +- 2 files changed, 180 insertions(+), 2412 deletions(-) diff --git a/Project/Tools/emoji tester.ipynb b/Project/Tools/emoji tester.ipynb index 1f3ad70..23f3462 100644 --- a/Project/Tools/emoji tester.ipynb +++ b/Project/Tools/emoji tester.ipynb @@ -7,7 +7,45 @@ "outputs": [], "source": [ "import pandas as pd\n", - "from IPython.display import clear_output, Markdown, Math" + "from IPython.display import clear_output, Markdown, Math\n", + "import ipywidgets as widgets\n", + "import os" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "----\n", + "## file input stuff:\n", + "\n", + "* replace `test.txt` with your whatsapp log file" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "================================================================================\n", + "processing File: test.txt\n", + "================================================================================\n", + "================================================================================\n", + "successfully finished action: processing File: test.txt\n", + "================================================================================\n", + "================================================================================\n", + "Wrote output to test.txt.csv\n", + "================================================================================\n" + ] + } + ], + "source": [ + "%%bash\n", + "./whatsapp2csv.sh test.txt" ] }, { @@ -19,235 +57,12 @@ }, { "cell_type": "code", - "execution_count": 2, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
datetimepersonmessage
010/18/1611:18Lena CartaHallo Jonas ๐Ÿ˜‰
110/18/1611:21Jonas WeinzMoin moin ;) Vllt. bin ich aber gar nicht Jona...
210/18/1611:21Lena Carta๐Ÿ˜ฑ๐Ÿค•
310/18/1611:21Sonja Vorwerk๐Ÿ™ˆ๐Ÿ˜‚
410/18/1611:22Caner Ucar๐Ÿ˜‚๐Ÿ˜‚๐Ÿ˜‚
510/18/1611:22Patrick (Arithmeum)hallo jeff
610/18/1611:22Caner UcarPatrick kein Handy am Empfang
710/18/1611:26Patrick (Arithmeum)halts maul und geh ma duschen
810/18/1611:34Simon (Arithmeum)Hab dich als Carsten abgespeichert, selber Schuld
910/18/1611:36Caner UcarPatrick ich hab geduscht mit Kristina und Sonja v
1010/18/1611:36Lena Cartaรคhm....
1110/18/1611:37Jonathan (Arithmeum)Caner benutzt das Damen-Bad
1210/18/1611:37Jonas WeinzLรผg doch nicht, du tanzt die ganze Zeit aufm T...
1310/18/1611:37Jonathan (Arithmeum)bรครคh๐Ÿ˜ฑ
1410/18/1611:40Patrick (Arithmeum)ja caner so ist das unter frauen..da geht man ...
1510/18/1612:02Thomas KoppelbergEeemmm was?
1610/18/1612:12Kristina StellwagThomas, du wolltest gestern ja nicht ๐Ÿ˜œ
1710/18/1612:14Thomas KoppelbergIch wollte einiges, aber ich kann mich zurรผckh...
1810/18/1612:19Simon (Arithmeum)Genau ๐Ÿ™Š
1910/18/1613:36Caner UcarSex
\n", - "
" - ], - "text/plain": [ - " date time person \\\n", - "0 10/18/16 11:18 Lena Carta \n", - "1 10/18/16 11:21 Jonas Weinz \n", - "2 10/18/16 11:21 Lena Carta \n", - "3 10/18/16 11:21 Sonja Vorwerk \n", - "4 10/18/16 11:22 Caner Ucar \n", - "5 10/18/16 11:22 Patrick (Arithmeum) \n", - "6 10/18/16 11:22 Caner Ucar \n", - "7 10/18/16 11:26 Patrick (Arithmeum) \n", - "8 10/18/16 11:34 Simon (Arithmeum) \n", - "9 10/18/16 11:36 Caner Ucar \n", - "10 10/18/16 11:36 Lena Carta \n", - "11 10/18/16 11:37 Jonathan (Arithmeum) \n", - "12 10/18/16 11:37 Jonas Weinz \n", - "13 10/18/16 11:37 Jonathan (Arithmeum) \n", - "14 10/18/16 11:40 Patrick (Arithmeum) \n", - "15 10/18/16 12:02 Thomas Koppelberg \n", - "16 10/18/16 12:12 Kristina Stellwag \n", - "17 10/18/16 12:14 Thomas Koppelberg \n", - "18 10/18/16 12:19 Simon (Arithmeum) \n", - "19 10/18/16 13:36 Caner Ucar \n", - "\n", - " message \n", - "0 Hallo Jonas ๐Ÿ˜‰ \n", - "1 Moin moin ;) Vllt. bin ich aber gar nicht Jona... \n", - "2 ๐Ÿ˜ฑ๐Ÿค• \n", - "3 ๐Ÿ™ˆ๐Ÿ˜‚ \n", - "4 ๐Ÿ˜‚๐Ÿ˜‚๐Ÿ˜‚ \n", - "5 hallo jeff \n", - "6 Patrick kein Handy am Empfang \n", - "7 halts maul und geh ma duschen \n", - "8 Hab dich als Carsten abgespeichert, selber Schuld \n", - "9 Patrick ich hab geduscht mit Kristina und Sonja v \n", - "10 รคhm.... \n", - "11 Caner benutzt das Damen-Bad \n", - "12 Lรผg doch nicht, du tanzt die ganze Zeit aufm T... \n", - "13 bรครคh๐Ÿ˜ฑ \n", - "14 ja caner so ist das unter frauen..da geht man ... \n", - "15 Eeemmm was? \n", - "16 Thomas, du wolltest gestern ja nicht ๐Ÿ˜œ \n", - "17 Ich wollte einiges, aber ich kann mich zurรผckh... \n", - "18 Genau ๐Ÿ™Š \n", - "19 Sex " - ] - }, - "execution_count": 2, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ - "messages = pd.read_csv('test.csv', delimiter='\\t')\n", - "messages[:20]" + "messages = pd.read_csv('test.txt.csv', delimiter='\\t')\n", + "messages.head()" ] }, { @@ -259,7 +74,31 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "found existing emoji specification\n" + ] + } + ], + "source": [ + "%%bash\n", + "if [ ! -e emoji-data.txt ]\n", + "then\n", + " echo \"downloading emoji specification\"\n", + " wget https://www.unicode.org/Public/emoji/11.0/emoji-data.txt\n", + "else\n", + " echo \"found existing emoji specification\"\n", + "fi\n" + ] + }, + { + "cell_type": "code", + "execution_count": 5, "metadata": {}, "outputs": [], "source": [ @@ -268,428 +107,21 @@ " 0x1F3FC,\n", " 0x1F3FD,\n", " 0x1F3FE,\n", - " 0x1F3FF\n", + " 0x1F3FF,\n", + " 0x2642,\n", + " 0x2640\n", "])" ] }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 6, "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
unicodetype
228231A..231BEmoji_Presentation
22923E9..23ECEmoji_Presentation
23023F0Emoji_Presentation
23123F3Emoji_Presentation
23225FD..25FEEmoji_Presentation
2332614..2615Emoji_Presentation
2342648..2653Emoji_Presentation
235267FEmoji_Presentation
2362693Emoji_Presentation
23726A1Emoji_Presentation
23826AA..26ABEmoji_Presentation
23926BD..26BEEmoji_Presentation
24026C4..26C5Emoji_Presentation
24126CEEmoji_Presentation
24226D4Emoji_Presentation
24326EAEmoji_Presentation
24426F2..26F3Emoji_Presentation
24526F5Emoji_Presentation
24626FAEmoji_Presentation
24726FDEmoji_Presentation
2482705Emoji_Presentation
249270A..270BEmoji_Presentation
2502728Emoji_Presentation
251274CEmoji_Presentation
252274EEmoji_Presentation
2532753..2755Emoji_Presentation
2542757Emoji_Presentation
2552795..2797Emoji_Presentation
25627B0Emoji_Presentation
25727BFEmoji_Presentation
.........
3291F6F7..1F6F8Emoji_Presentation
3301F6F9Emoji_Presentation
3311F910..1F918Emoji_Presentation
3321F919..1F91EEmoji_Presentation
3331F91FEmoji_Presentation
3341F920..1F927Emoji_Presentation
3351F928..1F92FEmoji_Presentation
3361F930Emoji_Presentation
3371F931..1F932Emoji_Presentation
3381F933..1F93AEmoji_Presentation
3391F93C..1F93EEmoji_Presentation
3401F940..1F945Emoji_Presentation
3411F947..1F94BEmoji_Presentation
3421F94CEmoji_Presentation
3431F94D..1F94FEmoji_Presentation
3441F950..1F95EEmoji_Presentation
3451F95F..1F96BEmoji_Presentation
3461F96C..1F970Emoji_Presentation
3471F973..1F976Emoji_Presentation
3481F97AEmoji_Presentation
3491F97C..1F97FEmoji_Presentation
3501F980..1F984Emoji_Presentation
3511F985..1F991Emoji_Presentation
3521F992..1F997Emoji_Presentation
3531F998..1F9A2Emoji_Presentation
3541F9B0..1F9B9Emoji_Presentation
3551F9C0Emoji_Presentation
3561F9C1..1F9C2Emoji_Presentation
3571F9D0..1F9E6Emoji_Presentation
3581F9E7..1F9FFEmoji_Presentation
\n", - "

131 rows ร— 2 columns

\n", - "
" - ], - "text/plain": [ - " unicode type\n", - "228 231A..231B Emoji_Presentation\n", - "229 23E9..23EC Emoji_Presentation\n", - "230 23F0 Emoji_Presentation\n", - "231 23F3 Emoji_Presentation\n", - "232 25FD..25FE Emoji_Presentation\n", - "233 2614..2615 Emoji_Presentation\n", - "234 2648..2653 Emoji_Presentation\n", - "235 267F Emoji_Presentation\n", - "236 2693 Emoji_Presentation\n", - "237 26A1 Emoji_Presentation\n", - "238 26AA..26AB Emoji_Presentation\n", - "239 26BD..26BE Emoji_Presentation\n", - "240 26C4..26C5 Emoji_Presentation\n", - "241 26CE Emoji_Presentation\n", - "242 26D4 Emoji_Presentation\n", - "243 26EA Emoji_Presentation\n", - "244 26F2..26F3 Emoji_Presentation\n", - "245 26F5 Emoji_Presentation\n", - "246 26FA Emoji_Presentation\n", - "247 26FD Emoji_Presentation\n", - "248 2705 Emoji_Presentation\n", - "249 270A..270B Emoji_Presentation\n", - "250 2728 Emoji_Presentation\n", - "251 274C Emoji_Presentation\n", - "252 274E Emoji_Presentation\n", - "253 2753..2755 Emoji_Presentation\n", - "254 2757 Emoji_Presentation\n", - "255 2795..2797 Emoji_Presentation\n", - "256 27B0 Emoji_Presentation\n", - "257 27BF Emoji_Presentation\n", - ".. ... ...\n", - "329 1F6F7..1F6F8 Emoji_Presentation\n", - "330 1F6F9 Emoji_Presentation\n", - "331 1F910..1F918 Emoji_Presentation\n", - "332 1F919..1F91E Emoji_Presentation\n", - "333 1F91F Emoji_Presentation\n", - "334 1F920..1F927 Emoji_Presentation\n", - "335 1F928..1F92F Emoji_Presentation\n", - "336 1F930 Emoji_Presentation\n", - "337 1F931..1F932 Emoji_Presentation\n", - "338 1F933..1F93A Emoji_Presentation\n", - "339 1F93C..1F93E Emoji_Presentation\n", - "340 1F940..1F945 Emoji_Presentation\n", - "341 1F947..1F94B Emoji_Presentation\n", - "342 1F94C Emoji_Presentation\n", - "343 1F94D..1F94F Emoji_Presentation\n", - "344 1F950..1F95E Emoji_Presentation\n", - "345 1F95F..1F96B Emoji_Presentation\n", - "346 1F96C..1F970 Emoji_Presentation\n", - "347 1F973..1F976 Emoji_Presentation\n", - "348 1F97A Emoji_Presentation\n", - "349 1F97C..1F97F Emoji_Presentation\n", - "350 1F980..1F984 Emoji_Presentation\n", - "351 1F985..1F991 Emoji_Presentation\n", - "352 1F992..1F997 Emoji_Presentation\n", - "353 1F998..1F9A2 Emoji_Presentation\n", - "354 1F9B0..1F9B9 Emoji_Presentation\n", - "355 1F9C0 Emoji_Presentation\n", - "356 1F9C1..1F9C2 Emoji_Presentation\n", - "357 1F9D0..1F9E6 Emoji_Presentation\n", - "358 1F9E7..1F9FF Emoji_Presentation\n", - "\n", - "[131 rows x 2 columns]" - ] - }, - "execution_count": 4, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ "emoji_data = pd.read_csv('emoji-data.txt', delimiter=';', comment='#', names=[\"unicode\",\"type\"])\n", "emoji_data['type'] = emoji_data['type'].str.strip()\n", - "emoji_data = emoji_data[emoji_data['type'] == \"Emoji_Presentation\"]\n", - "emoji_data" + "emoji_data = emoji_data[emoji_data['type'] == \"Emoji_Presentation\"]" ] }, { @@ -701,7 +133,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 7, "metadata": {}, "outputs": [ { @@ -710,7 +142,7 @@ "True" ] }, - "execution_count": 5, + "execution_count": 7, "metadata": {}, "output_type": "execute_result" } @@ -721,7 +153,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 8, "metadata": {}, "outputs": [ { @@ -735,7 +167,7 @@ "Name: unicode, dtype: object" ] }, - "execution_count": 6, + "execution_count": 8, "metadata": {}, "output_type": "execute_result" } @@ -754,9 +186,52 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 9, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "text/markdown": [ + "**used Emojis:**" + ], + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": [ + "'๐Ÿ€„\\U0001f9f1\\U0001f9f2\\U0001f9f3\\U0001f9f4\\U0001f9f5\\U0001f9f6๐Ÿƒ๐Ÿค๐Ÿค‘๐Ÿค’๐Ÿค“๐Ÿค”๐Ÿค•๐Ÿค–๐Ÿค—๐Ÿค˜๐Ÿค™๐Ÿคš๐Ÿค›๐Ÿคœ๐Ÿค๐Ÿคž\\U0001f91f๐Ÿค ๐Ÿคก๐Ÿคข๐Ÿคฃ๐Ÿคค๐Ÿคฅ๐Ÿคฆ๐Ÿคง\\U0001f928\\U0001f929\\U0001f92a\\U0001f92b\\U0001f92c\\U0001f92d\\U0001f92e\\U0001f92f๐Ÿคฐ\\U0001f931\\U0001f932๐Ÿคณ๐Ÿคด๐Ÿคต๐Ÿคถ๐Ÿคท๐Ÿคธ๐Ÿคน๐Ÿคบ๐Ÿคผ๐Ÿคฝ๐Ÿคพ๐Ÿฅ€๐Ÿฅ๐Ÿฅ‚๐Ÿฅƒ๐Ÿฅ„๐Ÿฅ…๐Ÿฅ‡๐Ÿฅˆ๐Ÿฅ‰๐ŸฅŠ๐Ÿฅ‹\\U0001f94c\\U0001f94d\\U0001f94e\\U0001f94f๐Ÿฅ๐Ÿฅ‘๐Ÿฅ’๐Ÿฅ“๐Ÿฅ”๐Ÿฅ•๐Ÿฅ–๐Ÿฅ—๐Ÿฅ˜๐Ÿฅ™๐Ÿฅš๐Ÿฅ›๐Ÿฅœ๐Ÿฅ๐Ÿฅž\\U0001f95f\\U0001f960\\U0001f961\\U0001f962\\U0001f963\\U0001f964\\U0001f965\\U0001f966\\U0001f967\\U0001f968\\U0001f969\\U0001f96a\\U0001f96b\\U0001f96c\\U0001f96d\\U0001f96e\\U0001f96f\\U0001f970\\U0001f9ec\\U0001f9ed\\U0001f973\\U0001f974\\U0001f975\\U0001f976\\U0001f9ee\\U0001f97a\\U0001f9ef\\U0001f97c\\U0001f97d\\U0001f97e\\U0001f97f๐Ÿฆ€๐Ÿฆ๐Ÿฆ‚๐Ÿฆƒ๐Ÿฆ„๐Ÿฆ…๐Ÿฆ†๐Ÿฆ‡๐Ÿฆˆ๐Ÿฆ‰๐ŸฆŠ๐Ÿฆ‹๐ŸฆŒ๐Ÿฆ๐Ÿ†Ž๐ŸฆŽ๐Ÿฆ๐Ÿ†‘๐Ÿ†’๐Ÿ†“๐Ÿ†”๐Ÿ†•๐Ÿ†–๐Ÿ†—๐Ÿ†˜๐Ÿ†™๐Ÿ†š\\U0001f992\\U0001f993\\U0001f994\\U0001f995\\U0001f996\\U0001f997\\U0001f998\\U0001f999\\U0001f99a\\U0001f99b\\U0001f99c\\U0001f99d\\U0001f99e\\U0001f99f\\U0001f9a0\\U0001f9a1\\U0001f9a2\\U0001f9f8\\U0001f9b0\\U0001f9b1\\U0001f9b2\\U0001f9b3\\U0001f9b4\\U0001f9b5\\U0001f9b6\\U0001f9b7\\U0001f9b8\\U0001f9b9๐Ÿง€\\U0001f9c1\\U0001f9c2\\U0001f9d0\\U0001f9d1\\U0001f9d2\\U0001f9d3\\U0001f9d4\\U0001f9d5\\U0001f9d6\\U0001f9d7\\U0001f9d8\\U0001f9d9\\U0001f9da\\U0001f9db\\U0001f9dc\\U0001f9dd\\U0001f9de\\U0001f9df\\U0001f9e0\\U0001f9e1\\U0001f9e2\\U0001f9e3\\U0001f9e4\\U0001f9e5๐Ÿ‡ฆ๐Ÿ‡ง๐Ÿ‡จ๐Ÿ‡ฉ๐Ÿ‡ช๐Ÿ‡ซ๐Ÿ‡ฌ๐Ÿ‡ญ๐Ÿ‡ฎ๐Ÿ‡ฏ๐Ÿ‡ฐ๐Ÿ‡ฑ๐Ÿ‡ฒ๐Ÿ‡ณ๐Ÿ‡ด๐Ÿ‡ต๐Ÿ‡ถ๐Ÿ‡ท๐Ÿ‡ธ๐Ÿ‡น๐Ÿ‡บ๐Ÿ‡ป๐Ÿ‡ผ๐Ÿ‡ฝ๐Ÿ‡พ๐Ÿ‡ฟ\\U0001f9f7๐Ÿˆ\\U0001f9f9\\U0001f9fa\\U0001f9fb\\U0001f9fc\\U0001f9fd\\U0001f9fe\\U0001f9ff๐Ÿˆš๐Ÿˆฏ๐Ÿˆฒ๐Ÿˆณ๐Ÿˆด๐Ÿˆต๐Ÿˆถ๐Ÿˆธ๐Ÿˆน๐Ÿˆบ๐Ÿ‰๐Ÿ‰‘๐ŸŒ€๐ŸŒ๐ŸŒ‚๐ŸŒƒ๐ŸŒ„๐ŸŒ…๐ŸŒ†๐ŸŒ‡๐ŸŒˆ๐ŸŒ‰๐ŸŒŠ๐ŸŒ‹๐ŸŒŒ๐ŸŒ๐ŸŒŽ๐ŸŒ๐ŸŒ๐ŸŒ‘๐ŸŒ’๐ŸŒ“๐ŸŒ”๐ŸŒ•๐ŸŒ–๐ŸŒ—๐ŸŒ˜๐ŸŒ™โŒšโŒ›โฌ›โฌœ๐ŸŒš๐ŸŒ›๐ŸŒœ๐ŸŒ๐ŸŒž๐ŸŒŸ๐ŸŒ ๐ŸŒญ๐ŸŒฎ๐ŸŒฏ๐ŸŒฐ๐ŸŒฑ๐ŸŒฒ๐ŸŒณ๐ŸŒด๐ŸŒต๐ŸŒท๐ŸŒธ๐ŸŒน๐ŸŒบ๐ŸŒป๐ŸŒผ๐ŸŒฝ๐ŸŒพ๐ŸŒฟ๐Ÿ€๐Ÿ๐Ÿ‚๐Ÿƒ๐Ÿ„๐Ÿ…๐Ÿ†๐Ÿ‡๐Ÿˆ๐Ÿ‰๐ŸŠ๐Ÿ‹๐ŸŒ๐Ÿ๐ŸŽ๐Ÿโญ๐Ÿ๐Ÿ‘๐Ÿ’๐Ÿ“โญ•๐Ÿ”๐Ÿ•๐Ÿ–๐Ÿ—๐Ÿ˜๐Ÿ™๐Ÿš๐Ÿ›๐Ÿœ๐ŸŸ๐Ÿ๐Ÿž๐Ÿ ๐Ÿก๐Ÿค๐Ÿข๐Ÿฃ๐Ÿฅ๐Ÿฆ๐Ÿฉ๐Ÿง๐Ÿจ๐Ÿช๐Ÿซ๐Ÿฎ๐Ÿฌ๐Ÿญ๐Ÿฏ๐Ÿฐ๐Ÿณ๐Ÿฑ๐Ÿฒ๐Ÿด๐Ÿต๐Ÿธ๐Ÿถ๐Ÿท๐Ÿน๐Ÿบ๐Ÿป๐Ÿผ๐Ÿพ๐Ÿฟ๐ŸŽ€๐ŸŽ๐ŸŽ‚๐ŸŽƒ๐ŸŽ…๐ŸŽ„๐ŸŽ†๐ŸŽˆ๐ŸŽ‰๐ŸŽŠ๐ŸŽ‹๐ŸŽŒ๐ŸŽ๐ŸŽ‡๐ŸŽ๐ŸŽ๐ŸŽŽ๐ŸŽ‘๐ŸŽ’๐ŸŽ“๐ŸŽ ๐ŸŽก๐ŸŽข๐ŸŽฃ๐ŸŽค๐ŸŽฅ๐ŸŽฆ๐ŸŽง๐ŸŽจ๐ŸŽฉ๐ŸŽช๐ŸŽซ๐ŸŽฌ๐ŸŽญ๐ŸŽฎ๐ŸŽฏ๐ŸŽฐ๐ŸŽฑ๐ŸŽฒ๐ŸŽณ๐ŸŽด๐ŸŽต๐ŸŽถ๐ŸŽท๐ŸŽธ๐ŸŽน๐ŸŽบ๐ŸŽป๐ŸŽผ๐ŸŽฝ๐ŸŽพ๐ŸŽฟ๐Ÿ€๐Ÿ๐Ÿ‚๐Ÿƒ๐Ÿ„๐Ÿ…๐Ÿ†๐Ÿ‡๐Ÿˆ๐Ÿ‰๐ŸŠ๐Ÿ๐Ÿ๐Ÿ‘๐Ÿ’๐Ÿ“๐Ÿ ๐Ÿก๐Ÿข๐Ÿฃ๐Ÿค๐Ÿฅ๐Ÿฆ๐Ÿง๐Ÿจ๐Ÿฉ๐Ÿช๐Ÿซ๐Ÿฌ๐Ÿญ๐Ÿฎ๐Ÿฏ๐Ÿฐโฉโชโซ๐Ÿดโฌโฐโณ๐Ÿธ๐Ÿน๐Ÿบ๐Ÿ€๐Ÿ๐Ÿ‚๐Ÿƒ๐Ÿ„๐Ÿ…๐Ÿ†๐Ÿ‡๐Ÿˆ๐Ÿ‰๐ŸŠ๐Ÿ‹๐ŸŒ๐Ÿ๐ŸŽ๐Ÿ๐Ÿ๐Ÿ‘๐Ÿ’๐Ÿ“๐Ÿ”๐Ÿ•๐Ÿ–๐Ÿ—๐Ÿ˜๐Ÿ™๐Ÿš๐Ÿ›๐Ÿœ๐Ÿ๐Ÿž๐ŸŸ๐Ÿ ๐Ÿก๐Ÿข๐Ÿฃ๐Ÿค๐Ÿฅ๐Ÿฆ๐Ÿง๐Ÿจ๐Ÿฉ๐Ÿช๐Ÿซ๐Ÿฌ๐Ÿญ๐Ÿฎ๐Ÿฏ๐Ÿฐ๐Ÿฑ๐Ÿฒ๐Ÿณ๐Ÿด๐Ÿต๐Ÿถ๐Ÿท๐Ÿธ๐Ÿน๐Ÿบ๐Ÿป๐Ÿผ๐Ÿฝ๐Ÿพ๐Ÿ‘€๐Ÿ‘‚๐Ÿ‘ƒ๐Ÿ‘„๐Ÿ‘…๐Ÿ‘†๐Ÿ‘‡๐Ÿ‘ˆ๐Ÿ‘‰๐Ÿ‘Š๐Ÿ‘‹๐Ÿ‘Œ๐Ÿ‘๐Ÿ‘Ž๐Ÿ‘๐Ÿ‘๐Ÿ‘‘๐Ÿ‘’๐Ÿ‘“๐Ÿ‘”๐Ÿ‘•๐Ÿ‘–๐Ÿ‘—๐Ÿ‘˜๐Ÿ‘™๐Ÿ‘š๐Ÿ‘›๐Ÿ‘œ๐Ÿ‘๐Ÿ‘ž๐Ÿ‘Ÿ๐Ÿ‘ ๐Ÿ‘ก๐Ÿ‘ข๐Ÿ‘ฃ๐Ÿ‘ค๐Ÿ‘ฅ๐Ÿ‘ฆ๐Ÿ‘ง๐Ÿ‘จ๐Ÿ‘ฉ๐Ÿ‘ช๐Ÿ‘ซ๐Ÿ‘ฌ๐Ÿ‘ญ๐Ÿ‘ฎ๐Ÿ‘ฏ๐Ÿ‘ฐ๐Ÿ‘ฑ๐Ÿ‘ฒ๐Ÿ‘ณ๐Ÿ‘ด๐Ÿ‘ต๐Ÿ‘ถ๐Ÿ‘ท๐Ÿ‘ธ๐Ÿ‘น๐Ÿ‘บ๐Ÿ‘ป๐Ÿ‘ผ๐Ÿ‘ฝ๐Ÿ‘พ๐Ÿ‘ฟ๐Ÿ’€๐Ÿ’๐Ÿ’‚๐Ÿ’ƒ๐Ÿ’„๐Ÿ’…๐Ÿ’†๐Ÿ’‡๐Ÿ’ˆ๐Ÿ’‰๐Ÿ’Š๐Ÿ’‹๐Ÿ’Œ๐Ÿ’๐Ÿ’Ž๐Ÿ’๐Ÿ’๐Ÿ’‘๐Ÿ’’๐Ÿ’“๐Ÿ’”๐Ÿ’•๐Ÿ’–๐Ÿ’—๐Ÿ’˜๐Ÿ’™๐Ÿ’š๐Ÿ’›๐Ÿ’œ๐Ÿ’๐Ÿ’ž๐Ÿ’Ÿ๐Ÿ’ ๐Ÿ’ก๐Ÿ’ข๐Ÿ’ฃ๐Ÿ’ค๐Ÿ’ฅ๐Ÿ’ฆ๐Ÿ’ง๐Ÿ’จ๐Ÿ’ฉ๐Ÿ’ช๐Ÿ’ซ๐Ÿ’ฌ๐Ÿ’ญ๐Ÿ’ฎ๐Ÿ’ฏ๐Ÿ’ฐ๐Ÿ’ฑ๐Ÿ’ฒ๐Ÿ’ณ๐Ÿ’ด๐Ÿ’ต๐Ÿ’ถ๐Ÿ’ท๐Ÿ’ธ๐Ÿ’น๐Ÿ’บ๐Ÿ’ป๐Ÿ’ผ๐Ÿ’ฝ๐Ÿ’พ๐Ÿ’ฟ๐Ÿ“€๐Ÿ“๐Ÿ“‚๐Ÿ“ƒ๐Ÿ“„๐Ÿ“…๐Ÿ“†๐Ÿ“‡๐Ÿ“ˆ๐Ÿ“‰๐Ÿ“Š๐Ÿ“‹๐Ÿ“Œ๐Ÿ“๐Ÿ“Ž๐Ÿ“๐Ÿ“๐Ÿ“‘๐Ÿ“’๐Ÿ““๐Ÿ“”๐Ÿ“•๐Ÿ“–๐Ÿ“—๐Ÿ“˜๐Ÿ“™๐Ÿ“š๐Ÿ“›๐Ÿ“œ๐Ÿ“๐Ÿ“ž๐Ÿ“Ÿ๐Ÿ“ ๐Ÿ“ก๐Ÿ“ข๐Ÿ“ฃ๐Ÿ“ค๐Ÿ“ฅ๐Ÿ“ฆ๐Ÿ“ง๐Ÿ“จ๐Ÿ“ฉ๐Ÿ“ช๐Ÿ“ซ๐Ÿ“ฌ๐Ÿ“ญ๐Ÿ“ฎ๐Ÿ“ฏ๐Ÿ“ฐ๐Ÿ“ฑ๐Ÿ“ฒ๐Ÿ“ณ๐Ÿ“ด๐Ÿ“ต๐Ÿ“ถ๐Ÿ“ท๐Ÿ“ธ๐Ÿ“น๐Ÿ“บ๐Ÿ“ป๐Ÿ“ผ๐Ÿ“ฟ๐Ÿ”€๐Ÿ”๐Ÿ”‚๐Ÿ”ƒ๐Ÿ”„๐Ÿ”…๐Ÿ”†๐Ÿ”‡๐Ÿ”ˆ๐Ÿ”‰๐Ÿ”Š๐Ÿ”‹๐Ÿ”Œ๐Ÿ”๐Ÿ”Ž๐Ÿ”๐Ÿ”๐Ÿ”‘๐Ÿ”’๐Ÿ”“๐Ÿ””๐Ÿ”•๐Ÿ”–๐Ÿ”—๐Ÿ”˜๐Ÿ”™๐Ÿ”š๐Ÿ”›๐Ÿ”œ๐Ÿ”๐Ÿ”ž๐Ÿ”Ÿ๐Ÿ” ๐Ÿ”ก๐Ÿ”ข๐Ÿ”ฃ๐Ÿ”ค๐Ÿ”ฅ๐Ÿ”ฆ๐Ÿ”ง๐Ÿ”จ๐Ÿ”ฉ๐Ÿ”ช๐Ÿ”ซ๐Ÿ”ฌ๐Ÿ”ญ๐Ÿ”ฎ๐Ÿ”ฏ๐Ÿ”ฐ๐Ÿ”ฑ๐Ÿ”ฒ๐Ÿ”ณ๐Ÿ”ด๐Ÿ”ต๐Ÿ”ถ๐Ÿ”ท๐Ÿ”ธ๐Ÿ”น๐Ÿ”บ๐Ÿ”ป๐Ÿ”ผ๐Ÿ”ฝ๐Ÿ•‹๐Ÿ•Œ๐Ÿ•๐Ÿ•Ž๐Ÿ•๐Ÿ•‘๐Ÿ•’๐Ÿ•“๐Ÿ•”๐Ÿ••๐Ÿ•–๐Ÿ•—๐Ÿ•˜๐Ÿ•™๐Ÿ•š๐Ÿ•›๐Ÿ•œ๐Ÿ•๐Ÿ•ž๐Ÿ•Ÿ๐Ÿ• ๐Ÿ•ก๐Ÿ•ข๐Ÿ•ฃ๐Ÿ•ค๐Ÿ•ฅ๐Ÿ•ฆ๐Ÿ•ง๐Ÿ•บ๐Ÿ–•๐Ÿ––๐Ÿ–ค๐Ÿ—ป๐Ÿ—ผโ—ฝโ—พ๐Ÿ—ฝ๐Ÿ—พ๐Ÿ—ฟ๐Ÿ˜€๐Ÿ˜๐Ÿ˜‚๐Ÿ˜ƒ๐Ÿ˜„๐Ÿ˜…๐Ÿ˜†๐Ÿ˜‡๐Ÿ˜ˆ๐Ÿ˜‰๐Ÿ˜Š๐Ÿ˜‹๐Ÿ˜Œ๐Ÿ˜๐Ÿ˜Ž๐Ÿ˜๐Ÿ˜๐Ÿ˜‘โ˜”โ˜•๐Ÿ˜’๐Ÿ˜“๐Ÿ˜”๐Ÿ˜•๐Ÿ˜–๐Ÿ˜—๐Ÿ˜˜๐Ÿ˜™๐Ÿ˜š๐Ÿ˜›๐Ÿ˜œ๐Ÿ˜๐Ÿ˜ž๐Ÿ˜Ÿ๐Ÿ˜ ๐Ÿ˜ก๐Ÿ˜ข๐Ÿ˜ฃ๐Ÿ˜ค๐Ÿ˜ฅ๐Ÿ˜ฆ๐Ÿ˜ง๐Ÿ˜จ๐Ÿ˜ฉ๐Ÿ˜ช๐Ÿ˜ซ๐Ÿ˜ฌ๐Ÿ˜ญ๐Ÿ˜ฎ๐Ÿ˜ฏ๐Ÿ˜ฐ๐Ÿ˜ฑ๐Ÿ˜ฒ๐Ÿ˜ณ๐Ÿ˜ด๐Ÿ˜ต๐Ÿ˜ถ๐Ÿ˜ท๐Ÿ˜ธ๐Ÿ˜น๐Ÿ˜บ๐Ÿ˜ป๐Ÿ˜ผ๐Ÿ˜ฝ๐Ÿ˜พ๐Ÿ˜ฟ๐Ÿ™€๐Ÿ™๐Ÿ™‚๐Ÿ™ƒโ™ˆโ™‰โ™Šโ™‹โ™Œโ™โ™Žโ™โ™โ™‘โ™’โ™“๐Ÿ™‹๐Ÿ™Œ๐Ÿ™๐Ÿ™Ž๐Ÿ™โ™ฟ๐Ÿš€๐Ÿš๐Ÿš‚๐Ÿšƒ๐Ÿš„๐Ÿš…๐Ÿš†๐Ÿš‡๐Ÿšˆ๐Ÿš‰๐ŸšŠ๐Ÿš‹๐ŸšŒ๐Ÿš๐ŸšŽ๐Ÿš๐Ÿš๐Ÿš‘๐Ÿš’โš“๐Ÿš“๐Ÿš”๐Ÿš•๐Ÿš–๐Ÿš—๐Ÿš˜๐Ÿš™๐Ÿšš๐Ÿš›๐Ÿšœ๐Ÿš๐Ÿšž๐ŸšŸโšก๐Ÿš ๐Ÿšก๐Ÿšข๐Ÿšฃ๐Ÿšค๐Ÿšฅ๐Ÿšฆ๐Ÿšงโšชโšซ๐Ÿšจ๐Ÿšฉ๐Ÿšช๐Ÿšซ๐Ÿšฌ๐Ÿšญ๐Ÿšฎ๐Ÿšฏ๐Ÿšฐ๐Ÿšฑ๐Ÿšฒ๐Ÿšณ๐Ÿšด๐Ÿšต๐Ÿšถ๐Ÿšท๐Ÿšธโšฝโšพ๐Ÿšน๐Ÿšบ๐Ÿšป๐Ÿšผ๐Ÿšฝโ›„โ›…๐Ÿšพ๐Ÿšฟ๐Ÿ›€๐Ÿ›๐Ÿ›‚๐Ÿ›ƒ๐Ÿ›„๐Ÿ›…โ›Ž๐Ÿ›Œ๐Ÿ›๐Ÿ›‘๐Ÿ›’โ›”โ›ช๐Ÿ›ซ๐Ÿ›ฌโ›ฒโ›ณ๐Ÿ›ดโ›ต๐Ÿ›ต๐Ÿ›ถ\\U0001f6f7\\U0001f6f8โ›บ\\U0001f6f9โ›ฝโœ…\\U0001f9e6๐Ÿ™„โœŠโœ‹๐Ÿ™…๐Ÿ™†๐Ÿ™‡๐Ÿ™ˆ\\U0001f9e7๐Ÿ™‰๐Ÿ™Šโœจ\\U0001f9e8โŒโŽ\\U0001f9e9โ“โ”โ•โ—\\U0001f9ea\\U0001f9ebโž•โž–โž—๐Ÿฆ๐Ÿฆ‘โžฐโžฟ\\U0001f9f0'" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/markdown": [ + "**blacklisted Emojis:**" + ], + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": [ + "'โ™€โ™‚๐Ÿป๐Ÿผ๐Ÿฝ๐Ÿพ๐Ÿฟ'" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], "source": [ "emoji_code_list = []\n", "for entry in emoji_codes:\n", @@ -771,12 +246,16 @@ " # single entry\n", " if i not in emoji_blacklist:\n", " emoji_code_list.append(int(entry,16))\n", - "emoji_code_set = set(emoji_code_list)" + "emoji_code_set = set(emoji_code_list)\n", + "display(Markdown(\"**used Emojis:**\"))\n", + "display(\"\".join([chr(x) for x in emoji_code_set]))\n", + "display(Markdown(\"**blacklisted Emojis:**\"))\n", + "display(\"\".join([chr(x) for x in emoji_blacklist]))" ] }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 10, "metadata": {}, "outputs": [ { @@ -801,7 +280,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 11, "metadata": {}, "outputs": [], "source": [ @@ -810,276 +289,30 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "for i in messages.index:\n", " emoji_list = []\n", - " to_remove = []\n", " m = messages.iloc[i]['message']\n", + " m_new = \"\"\n", " for c in str(m):\n", " if ord(c) in emoji_code_set:\n", " emoji_list.append(c)\n", - " elif ord(c) in emoji_blacklist:\n", - " to_remove.append(c)\n", + " elif ord(c) not in emoji_blacklist:\n", + " m_new += c\n", " \n", - " messages.loc[i,'emojis'] = emoji_list\n", + " messages.loc[i,'emojis'] = set(emoji_list)\n", " #remove emiรณjis from message\n", - " for e in (emoji_list + to_remove):\n", - " m = m.replace(e,\"\")\n", - " messages.loc[i,'message'] = m\n" + " messages.loc[i,'message'] = m_new\n" ] }, { "cell_type": "code", - "execution_count": 11, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
datetimepersonmessageemojis
010/18/1611:18Lena CartaHallo Jonas[๐Ÿ˜‰]
110/18/1611:21Jonas WeinzMoin moin ;) Vllt. bin ich aber gar nicht Jona...[๐Ÿค”]
210/18/1611:21Lena Carta[๐Ÿ˜ฑ, ๐Ÿค•]
310/18/1611:21Sonja Vorwerk[๐Ÿ™ˆ, ๐Ÿ˜‚]
410/18/1611:22Caner Ucar[๐Ÿ˜‚, ๐Ÿ˜‚, ๐Ÿ˜‚]
510/18/1611:22Patrick (Arithmeum)hallo jeff[]
610/18/1611:22Caner UcarPatrick kein Handy am Empfang[]
710/18/1611:26Patrick (Arithmeum)halts maul und geh ma duschen[]
810/18/1611:34Simon (Arithmeum)Hab dich als Carsten abgespeichert, selber Schuld[]
910/18/1611:36Caner UcarPatrick ich hab geduscht mit Kristina und Sonja v[]
1010/18/1611:36Lena Cartaรคhm....[]
1110/18/1611:37Jonathan (Arithmeum)Caner benutzt das Damen-Bad[]
1210/18/1611:37Jonas WeinzLรผg doch nicht, du tanzt die ganze Zeit aufm T...[]
1310/18/1611:37Jonathan (Arithmeum)bรครคh[๐Ÿ˜ฑ]
1410/18/1611:40Patrick (Arithmeum)ja caner so ist das unter frauen..da geht man ...[]
1510/18/1612:02Thomas KoppelbergEeemmm was?[]
1610/18/1612:12Kristina StellwagThomas, du wolltest gestern ja nicht[๐Ÿ˜œ]
1710/18/1612:14Thomas KoppelbergIch wollte einiges, aber ich kann mich zurรผckh...[]
1810/18/1612:19Simon (Arithmeum)Genau[๐Ÿ™Š]
1910/18/1613:36Caner UcarSex[]
\n", - "
" - ], - "text/plain": [ - " date time person \\\n", - "0 10/18/16 11:18 Lena Carta \n", - "1 10/18/16 11:21 Jonas Weinz \n", - "2 10/18/16 11:21 Lena Carta \n", - "3 10/18/16 11:21 Sonja Vorwerk \n", - "4 10/18/16 11:22 Caner Ucar \n", - "5 10/18/16 11:22 Patrick (Arithmeum) \n", - "6 10/18/16 11:22 Caner Ucar \n", - "7 10/18/16 11:26 Patrick (Arithmeum) \n", - "8 10/18/16 11:34 Simon (Arithmeum) \n", - "9 10/18/16 11:36 Caner Ucar \n", - "10 10/18/16 11:36 Lena Carta \n", - "11 10/18/16 11:37 Jonathan (Arithmeum) \n", - "12 10/18/16 11:37 Jonas Weinz \n", - "13 10/18/16 11:37 Jonathan (Arithmeum) \n", - "14 10/18/16 11:40 Patrick (Arithmeum) \n", - "15 10/18/16 12:02 Thomas Koppelberg \n", - "16 10/18/16 12:12 Kristina Stellwag \n", - "17 10/18/16 12:14 Thomas Koppelberg \n", - "18 10/18/16 12:19 Simon (Arithmeum) \n", - "19 10/18/16 13:36 Caner Ucar \n", - "\n", - " message emojis \n", - "0 Hallo Jonas [๐Ÿ˜‰] \n", - "1 Moin moin ;) Vllt. bin ich aber gar nicht Jona... [๐Ÿค”] \n", - "2 [๐Ÿ˜ฑ, ๐Ÿค•] \n", - "3 [๐Ÿ™ˆ, ๐Ÿ˜‚] \n", - "4 [๐Ÿ˜‚, ๐Ÿ˜‚, ๐Ÿ˜‚] \n", - "5 hallo jeff [] \n", - "6 Patrick kein Handy am Empfang [] \n", - "7 halts maul und geh ma duschen [] \n", - "8 Hab dich als Carsten abgespeichert, selber Schuld [] \n", - "9 Patrick ich hab geduscht mit Kristina und Sonja v [] \n", - "10 รคhm.... [] \n", - "11 Caner benutzt das Damen-Bad [] \n", - "12 Lรผg doch nicht, du tanzt die ganze Zeit aufm T... [] \n", - "13 bรครคh [๐Ÿ˜ฑ] \n", - "14 ja caner so ist das unter frauen..da geht man ... [] \n", - "15 Eeemmm was? [] \n", - "16 Thomas, du wolltest gestern ja nicht [๐Ÿ˜œ] \n", - "17 Ich wollte einiges, aber ich kann mich zurรผckh... [] \n", - "18 Genau [๐Ÿ™Š] \n", - "19 Sex [] " - ] - }, - "execution_count": 11, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ "messages[:20]" ] @@ -1093,7 +326,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 14, "metadata": {}, "outputs": [], "source": [ @@ -1103,728 +336,9 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
datetimepersonmessageemojis
010/18/1611:18Lena CartaHallo Jonas[๐Ÿ˜‰]
110/18/1611:21Jonas WeinzMoin moin ;) Vllt. bin ich aber gar nicht Jona...[๐Ÿค”]
1310/18/1611:37Jonathan (Arithmeum)bรครคh[๐Ÿ˜ฑ]
1610/18/1612:12Kristina StellwagThomas, du wolltest gestern ja nicht[๐Ÿ˜œ]
1810/18/1612:19Simon (Arithmeum)Genau[๐Ÿ™Š]
2910/18/1618:38Lena Cartawas du bist morgen nicht da ???[๐Ÿ˜ฑ, ๐Ÿ˜ฏ, ๐Ÿ˜”]
3110/18/1618:39Lena Cartaich dachte diese Woche komplett noch. Aber gut...[๐Ÿ˜‡]
3210/18/1618:40Kristina StellwagNee das Semester hat doch angefangen[๐Ÿ˜œ]
3310/18/1618:42Lena Cartaja hatte ich gerade vergessen . Dabei hatte ic...[๐Ÿ˜‚, ๐Ÿ˜ถ]
4110/20/1612:50Steffi (Arithmeum)Klar[๐Ÿ‘น]
4410/20/1617:54Steffi (Arithmeum)Mein Zug lรคsst sich auch heute Zeit[๐Ÿ˜ž]
4510/20/1618:08Steffi (Arithmeum)Wir fangen doch wahrscheinlich nicht vor halb ...[๐Ÿ˜ณ]
5110/20/1620:58Anna BorutzkyIst das dein Ernst?[๐Ÿ™„]
5210/20/1620:58Patrick (Arithmeum)Habs voll vercheckt sry ..gerade erst gemerkt[๐Ÿ™„]
7010/20/1622:59Jonathan (Arithmeum)Ich bitte darum Caner....[๐Ÿ˜, ๐Ÿ˜, ๐Ÿ˜]
7810/25/1601:03Caner UcarHaaaaaalloooo[๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿ˜, ๐Ÿค—, ๐Ÿค—, ๐Ÿ˜]
7910/25/1601:04Michael (Arithmeum)Keine sorge, das hat bei dir nichts mit der St...[๐Ÿ˜‰]
8810/25/1609:02Lena Cartaich wollte eigentlich schon da sein, aber ich ...[๐Ÿ˜‡]
8910/25/1609:04Jonathan (Arithmeum)SIIIIEG...โš”[๐ŸŽ‰]
9010/25/1610:00Lena Cartabesser spรคt als nie. ich bin immerhin auf dem ...[๐Ÿ˜]
9310/25/1621:49Jonathan (Arithmeum)Ich nicht[๐Ÿ˜„]
9410/25/1621:50Lena Cartaich auch nicht[๐Ÿ˜ณ]
9510/25/1622:13Konstantin (Arithmeum)Wer hat den zuletzt was laminiert? lena und i...[๐Ÿค”]
9610/25/1623:28Caner UcarIch auch nicht[๐Ÿค”]
10210/27/1614:43Sonja Vorwerkhm mรผsste das am besten heute schon wissen...w...[๐Ÿ™ˆ, ๐Ÿ˜„]
11510/28/1608:32Lena CartaGuten Morgen liebe Kollegen, kรถnnte einer von ...[๐Ÿ™]
11610/28/1609:48Lena CartaIst noch niemand im Bรผro?[๐Ÿ˜‡]
11710/28/1609:50Caner Ucar10:10 ca.[๐Ÿ™Œ]
11810/28/1609:51Caner UcarWann hat das woki geรถffnet[๐Ÿ˜„]
12110/28/1610:54Lena CartaDa hรคtte ich jetzt auch nicht gesucht[๐Ÿ˜]
..................
86635/9/1817:42Jonas Weinzmehrere Bรผrostรผhle hintereinander?[๐Ÿค”]
86655/9/1817:43Lena CartaWollte Kristina das nicht immer schonmal machen[๐Ÿ˜]
86755/10/1814:06Lena KlรถcknerErledigt[๐Ÿ‘]
86765/10/1814:07Caner UcarDanke[๐Ÿ˜Š]
86775/10/1814:23Anna ZensIch hatte es auch schon erledigt[๐Ÿ˜„]
86785/11/1808:16Lena CartaGute morgen, ich habe gesehen neben dem Buch l...[๐Ÿ™‚]
86795/11/1808:25Anna ZensDie handschriftlichen sind die, von denen kein...[๐Ÿ˜„]
86805/11/1808:26Lena Cartaachso ich denke mal dann wird sich Ingo melden[๐Ÿ˜€, ๐Ÿ˜€]
86815/11/1813:18Freya (Arithmeum)Mein Kalender funktioniert nicht - kรถnnte mir ...[๐Ÿค”]
86825/12/1814:53Freya (Arithmeum)Kรถnnte jemand fรผr Montag Abend zwei Karten fรผr...[๐Ÿ˜‡]
86835/12/1815:05Anna BorutzkyUnd eine Freikarte fรผr Sonja bitte![๐Ÿ˜€]
86895/12/1815:16Lena CartaDanke Wieso kommen solche Fragen immer am Woc...[๐Ÿ˜€]
86925/12/1815:17Anna BorutzkyMurphyโ€˜s Gesetz?[๐Ÿ˜‰]
86935/12/1815:17Caner UcarCool[๐Ÿ˜]
86955/12/1815:23Jonas Weinz...und die kommt auf Annas Tisch[๐Ÿ˜‰]
86965/12/1815:23Freya (Arithmeum)Vielen, vielen Dank โ˜บ Sowas passiert immer am ...[๐Ÿ™ˆ]
86975/12/1815:23Lena CartaDas war nicht auf dich bezogen .[๐Ÿ˜€]
86985/12/1815:24Freya (Arithmeum)Na gut[๐Ÿ˜„]
86995/12/1815:24Lena CartaEigentlich ist es halbwegs entspannt heute Ic...[๐Ÿ˜€, ๐Ÿ˜›]
87005/12/1815:25Freya (Arithmeum)Es gibt keine doofen Fragen[๐Ÿ˜‹]
87025/12/1815:25Lena Cartahttps://www.youtube.com/watch?v=uPHi5xn_q5c[๐Ÿ˜›]
87085/12/1817:56Freya (Arithmeum)Ich mรถchte arbeiten und habe mega Bock drauf! ...[๐Ÿ˜‡]
87195/12/1820:09Lena Cartabin leider verabredet .... mit Thilo[๐Ÿ˜ฌ, ๐Ÿค—]
87255/13/1817:24Thomas Koppelbergโ™ฅ[๐Ÿ˜]
87265/14/1817:00Sonja SchรคferKann mir jemand noch eine zweite Freikarte res...[๐Ÿ˜Š]
87395/14/1822:40Jonas WeinzHab Ingo beim vorbeilaufenden nicht gesehen, h...[๐Ÿ’]
87445/15/1817:16Caner Ucarโ€โ™€[๐Ÿคท]
87555/17/1810:43Lena Cartasuper perfekt dann ist beiden geholfen[๐Ÿ™‚, ๐Ÿ˜Ž]
87595/17/1810:43Lena Cartaob das so sinnvoll ist?[๐Ÿ˜€]
87635/17/1810:45Lena Cartaich dachte es ging konkret um den Termin[๐Ÿ˜€]
\n", - "

2455 rows ร— 5 columns

\n", - "
" - ], - "text/plain": [ - " date time person \\\n", - "0 10/18/16 11:18 Lena Carta \n", - "1 10/18/16 11:21 Jonas Weinz \n", - "13 10/18/16 11:37 Jonathan (Arithmeum) \n", - "16 10/18/16 12:12 Kristina Stellwag \n", - "18 10/18/16 12:19 Simon (Arithmeum) \n", - "29 10/18/16 18:38 Lena Carta \n", - "31 10/18/16 18:39 Lena Carta \n", - "32 10/18/16 18:40 Kristina Stellwag \n", - "33 10/18/16 18:42 Lena Carta \n", - "41 10/20/16 12:50 Steffi (Arithmeum) \n", - "44 10/20/16 17:54 Steffi (Arithmeum) \n", - "45 10/20/16 18:08 Steffi (Arithmeum) \n", - "51 10/20/16 20:58 Anna Borutzky \n", - "52 10/20/16 20:58 Patrick (Arithmeum) \n", - "70 10/20/16 22:59 Jonathan (Arithmeum) \n", - "78 10/25/16 01:03 Caner Ucar \n", - "79 10/25/16 01:04 Michael (Arithmeum) \n", - "88 10/25/16 09:02 Lena Carta \n", - "89 10/25/16 09:04 Jonathan (Arithmeum) \n", - "90 10/25/16 10:00 Lena Carta \n", - "93 10/25/16 21:49 Jonathan (Arithmeum) \n", - "94 10/25/16 21:50 Lena Carta \n", - "95 10/25/16 22:13 Konstantin (Arithmeum) \n", - "96 10/25/16 23:28 Caner Ucar \n", - "102 10/27/16 14:43 Sonja Vorwerk \n", - "115 10/28/16 08:32 Lena Carta \n", - "116 10/28/16 09:48 Lena Carta \n", - "117 10/28/16 09:50 Caner Ucar \n", - "118 10/28/16 09:51 Caner Ucar \n", - "121 10/28/16 10:54 Lena Carta \n", - "... ... ... ... \n", - "8663 5/9/18 17:42 Jonas Weinz \n", - "8665 5/9/18 17:43 Lena Carta \n", - "8675 5/10/18 14:06 Lena Klรถckner \n", - "8676 5/10/18 14:07 Caner Ucar \n", - "8677 5/10/18 14:23 Anna Zens \n", - "8678 5/11/18 08:16 Lena Carta \n", - "8679 5/11/18 08:25 Anna Zens \n", - "8680 5/11/18 08:26 Lena Carta \n", - "8681 5/11/18 13:18 Freya (Arithmeum) \n", - "8682 5/12/18 14:53 Freya (Arithmeum) \n", - "8683 5/12/18 15:05 Anna Borutzky \n", - "8689 5/12/18 15:16 Lena Carta \n", - "8692 5/12/18 15:17 Anna Borutzky \n", - "8693 5/12/18 15:17 Caner Ucar \n", - "8695 5/12/18 15:23 Jonas Weinz \n", - "8696 5/12/18 15:23 Freya (Arithmeum) \n", - "8697 5/12/18 15:23 Lena Carta \n", - "8698 5/12/18 15:24 Freya (Arithmeum) \n", - "8699 5/12/18 15:24 Lena Carta \n", - "8700 5/12/18 15:25 Freya (Arithmeum) \n", - "8702 5/12/18 15:25 Lena Carta \n", - "8708 5/12/18 17:56 Freya (Arithmeum) \n", - "8719 5/12/18 20:09 Lena Carta \n", - "8725 5/13/18 17:24 Thomas Koppelberg \n", - "8726 5/14/18 17:00 Sonja Schรคfer \n", - "8739 5/14/18 22:40 Jonas Weinz \n", - "8744 5/15/18 17:16 Caner Ucar \n", - "8755 5/17/18 10:43 Lena Carta \n", - "8759 5/17/18 10:43 Lena Carta \n", - "8763 5/17/18 10:45 Lena Carta \n", - "\n", - " message \\\n", - "0 Hallo Jonas \n", - "1 Moin moin ;) Vllt. bin ich aber gar nicht Jona... \n", - "13 bรครคh \n", - "16 Thomas, du wolltest gestern ja nicht \n", - "18 Genau \n", - "29 was du bist morgen nicht da ??? \n", - "31 ich dachte diese Woche komplett noch. Aber gut... \n", - "32 Nee das Semester hat doch angefangen \n", - "33 ja hatte ich gerade vergessen . Dabei hatte ic... \n", - "41 Klar \n", - "44 Mein Zug lรคsst sich auch heute Zeit \n", - "45 Wir fangen doch wahrscheinlich nicht vor halb ... \n", - "51 Ist das dein Ernst? \n", - "52 Habs voll vercheckt sry ..gerade erst gemerkt \n", - "70 Ich bitte darum Caner.... \n", - "78 Haaaaaalloooo \n", - "79 Keine sorge, das hat bei dir nichts mit der St... \n", - "88 ich wollte eigentlich schon da sein, aber ich ... \n", - "89 SIIIIEG...โš” \n", - "90 besser spรคt als nie. ich bin immerhin auf dem ... \n", - "93 Ich nicht \n", - "94 ich auch nicht \n", - "95 Wer hat den zuletzt was laminiert? lena und i... \n", - "96 Ich auch nicht \n", - "102 hm mรผsste das am besten heute schon wissen...w... \n", - "115 Guten Morgen liebe Kollegen, kรถnnte einer von ... \n", - "116 Ist noch niemand im Bรผro? \n", - "117 10:10 ca. \n", - "118 Wann hat das woki geรถffnet \n", - "121 Da hรคtte ich jetzt auch nicht gesucht \n", - "... ... \n", - "8663 mehrere Bรผrostรผhle hintereinander? \n", - "8665 Wollte Kristina das nicht immer schonmal machen \n", - "8675 Erledigt \n", - "8676 Danke \n", - "8677 Ich hatte es auch schon erledigt \n", - "8678 Gute morgen, ich habe gesehen neben dem Buch l... \n", - "8679 Die handschriftlichen sind die, von denen kein... \n", - "8680 achso ich denke mal dann wird sich Ingo melden \n", - "8681 Mein Kalender funktioniert nicht - kรถnnte mir ... \n", - "8682 Kรถnnte jemand fรผr Montag Abend zwei Karten fรผr... \n", - "8683 Und eine Freikarte fรผr Sonja bitte! \n", - "8689 Danke Wieso kommen solche Fragen immer am Woc... \n", - "8692 Murphyโ€˜s Gesetz? \n", - "8693 Cool \n", - "8695 ...und die kommt auf Annas Tisch \n", - "8696 Vielen, vielen Dank โ˜บ Sowas passiert immer am ... \n", - "8697 Das war nicht auf dich bezogen . \n", - "8698 Na gut \n", - "8699 Eigentlich ist es halbwegs entspannt heute Ic... \n", - "8700 Es gibt keine doofen Fragen \n", - "8702 https://www.youtube.com/watch?v=uPHi5xn_q5c \n", - "8708 Ich mรถchte arbeiten und habe mega Bock drauf! ... \n", - "8719 bin leider verabredet .... mit Thilo \n", - "8725 โ™ฅ \n", - "8726 Kann mir jemand noch eine zweite Freikarte res... \n", - "8739 Hab Ingo beim vorbeilaufenden nicht gesehen, h... \n", - "8744 โ€โ™€ \n", - "8755 super perfekt dann ist beiden geholfen \n", - "8759 ob das so sinnvoll ist? \n", - "8763 ich dachte es ging konkret um den Termin \n", - "\n", - " emojis \n", - "0 [๐Ÿ˜‰] \n", - "1 [๐Ÿค”] \n", - "13 [๐Ÿ˜ฑ] \n", - "16 [๐Ÿ˜œ] \n", - "18 [๐Ÿ™Š] \n", - "29 [๐Ÿ˜ฑ, ๐Ÿ˜ฏ, ๐Ÿ˜”] \n", - "31 [๐Ÿ˜‡] \n", - "32 [๐Ÿ˜œ] \n", - "33 [๐Ÿ˜‚, ๐Ÿ˜ถ] \n", - "41 [๐Ÿ‘น] \n", - "44 [๐Ÿ˜ž] \n", - "45 [๐Ÿ˜ณ] \n", - "51 [๐Ÿ™„] \n", - "52 [๐Ÿ™„] \n", - "70 [๐Ÿ˜, ๐Ÿ˜, ๐Ÿ˜] \n", - "78 [๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿค—, ๐Ÿ˜, ๐Ÿค—, ๐Ÿค—, ๐Ÿ˜] \n", - "79 [๐Ÿ˜‰] \n", - "88 [๐Ÿ˜‡] \n", - "89 [๐ŸŽ‰] \n", - "90 [๐Ÿ˜] \n", - "93 [๐Ÿ˜„] \n", - "94 [๐Ÿ˜ณ] \n", - "95 [๐Ÿค”] \n", - "96 [๐Ÿค”] \n", - "102 [๐Ÿ™ˆ, ๐Ÿ˜„] \n", - "115 [๐Ÿ™] \n", - "116 [๐Ÿ˜‡] \n", - "117 [๐Ÿ™Œ] \n", - "118 [๐Ÿ˜„] \n", - "121 [๐Ÿ˜] \n", - "... ... \n", - "8663 [๐Ÿค”] \n", - "8665 [๐Ÿ˜] \n", - "8675 [๐Ÿ‘] \n", - "8676 [๐Ÿ˜Š] \n", - "8677 [๐Ÿ˜„] \n", - "8678 [๐Ÿ™‚] \n", - "8679 [๐Ÿ˜„] \n", - "8680 [๐Ÿ˜€, ๐Ÿ˜€] \n", - "8681 [๐Ÿค”] \n", - "8682 [๐Ÿ˜‡] \n", - "8683 [๐Ÿ˜€] \n", - "8689 [๐Ÿ˜€] \n", - "8692 [๐Ÿ˜‰] \n", - "8693 [๐Ÿ˜] \n", - "8695 [๐Ÿ˜‰] \n", - "8696 [๐Ÿ™ˆ] \n", - "8697 [๐Ÿ˜€] \n", - "8698 [๐Ÿ˜„] \n", - "8699 [๐Ÿ˜€, ๐Ÿ˜›] \n", - "8700 [๐Ÿ˜‹] \n", - "8702 [๐Ÿ˜›] \n", - "8708 [๐Ÿ˜‡] \n", - "8719 [๐Ÿ˜ฌ, ๐Ÿค—] \n", - "8725 [๐Ÿ˜] \n", - "8726 [๐Ÿ˜Š] \n", - "8739 [๐Ÿ’] \n", - "8744 [๐Ÿคท] \n", - "8755 [๐Ÿ™‚, ๐Ÿ˜Ž] \n", - "8759 [๐Ÿ˜€] \n", - "8763 [๐Ÿ˜€] \n", - "\n", - "[2455 rows x 5 columns]" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], + "outputs": [], "source": [ "display(emoji_messages)" ] @@ -1839,7 +353,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 16, "metadata": {}, "outputs": [], "source": [ @@ -1847,36 +361,8 @@ "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" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [], - "source": [ - "labels=[e[-1] for e in emoji_messages['emojis']]" - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "['๐Ÿ˜‰', '๐Ÿค”', '๐Ÿ˜ฑ', '๐Ÿ˜œ', '๐Ÿ™Š', '๐Ÿ˜”', '๐Ÿ˜‡', '๐Ÿ˜œ', '๐Ÿ˜ถ', '๐Ÿ‘น']" - ] - }, - "execution_count": 16, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "labels[:10]" + "from sklearn.model_selection import train_test_split\n", + "from sklearn.preprocessing import MultiLabelBinarizer" ] }, { @@ -1884,13 +370,31 @@ "execution_count": 17, "metadata": {}, "outputs": [], + "source": [ + "mlb = MultiLabelBinarizer()\n", + "\n", + "labels=mlb.fit_transform(emoji_messages['emojis'])" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": {}, + "outputs": [], "source": [ "X1, Xt1, y1, yt1 = train_test_split(emoji_messages['message'], labels, test_size=0.1, random_state=4222)" ] }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 19, "metadata": {}, "outputs": [], "source": [ @@ -1901,782 +405,46 @@ }, { "cell_type": "code", - "execution_count": 19, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/home/jonas/.local/lib/python3.6/site-packages/sklearn/neural_network/multilayer_perceptron.py:564: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n", - " % self.max_iter, ConvergenceWarning)\n" - ] - }, - { - "data": { - "text/plain": [ - "MLPClassifier(activation='relu', alpha=0.0001, batch_size='auto', beta_1=0.9,\n", - " beta_2=0.999, early_stopping=False, epsilon=1e-08,\n", - " hidden_layer_sizes=(100,), learning_rate='constant',\n", - " learning_rate_init=0.001, max_iter=200, momentum=0.9,\n", - " nesterovs_momentum=True, power_t=0.5, random_state=None,\n", - " shuffle=True, solver='adam', tol=0.0001, validation_fraction=0.1,\n", - " verbose=False, warm_start=False)" - ] - }, - "execution_count": 19, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ "from sklearn.ensemble import RandomForestClassifier as RFC\n", "from sklearn.neural_network import MLPClassifier as MLP\n", "from sklearn.naive_bayes import MultinomialNB as MNB\n", - "#clf_a = RFC(criterion='entropy', random_state=4222)\n", - "clf_a = MLP()\n", + "from sklearn.multiclass import OneVsRestClassifier as OVRC\n", + "#clf_a = OVRC(RFC(criterion='entropy', random_state=4222))\n", + "clf_a = OVRC(MLP(hidden_layer_sizes=(10,)))\n", + "#clf_a = OVRC(MNB())\n", "clf_a.fit(vec_train, y1)" ] }, - { - "cell_type": "code", - "execution_count": 20, - "metadata": {}, - "outputs": [], - "source": [ - "pred = clf_a.predict(vectorizer.transform(Xt1))\n" - ] - }, - { - "cell_type": "code", - "execution_count": 21, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
messagepredtrainedemojis
8663mehrere Bรผrostรผhle hintereinander?๐Ÿ™ˆ๐Ÿค”[๐Ÿค”]
1862Ich will auch Essen๐Ÿ˜‰๐Ÿ˜ž[๐Ÿ˜ž]
39212x orange๐Ÿ™‚๐Ÿ™‚[๐Ÿ™‚]
2222Man soll auch keine Lernsachen da lassen๐Ÿ˜€๐Ÿ˜€[๐Ÿ˜€]
761Wir kรถnnen auch gerne wieder bei mir zusammen ...๐Ÿ˜‡๐Ÿ˜‰[๐Ÿ˜‰]
3600Hab den Bus verpasst, den ich eigentlich nehme...๐Ÿ˜Ÿ๐Ÿ‘ธ[๐Ÿ‘ธ]
1794Klaro, gerne๐Ÿ˜Ž๐Ÿ™‚[๐Ÿ™‚]
861Sehr gut, Jan! Wenn sich noch jemand melden wรผ...๐Ÿ˜˜๐Ÿ˜‡[๐Ÿ˜‡]
1469Jaaa aber mein Mitbewohner ist messy๐Ÿ˜ฌ๐Ÿ™Š[๐Ÿ˜ฑ, ๐Ÿ˜”, ๐ŸŒš, ๐Ÿ™Š]
6254Hallรถchen Es hat ein Herr angerufen, der gern...๐Ÿ’๐Ÿ™‚[๐Ÿ™‚]
7968Iwo๐Ÿคฆ๐Ÿ˜‚[๐Ÿ˜‚]
8102Keine Ahnung wer da haariger ist๐Ÿ˜…๐Ÿง[๐Ÿง]
6835Danke๐Ÿ˜Š๐Ÿ˜Š[๐Ÿ˜Š]
7135Wurde bestimmt von Big Bang geklaut๐Ÿคท๐Ÿ˜‚[๐Ÿ˜‚, ๐Ÿ˜‚, ๐Ÿ˜‚]
4556Ich kann leider nicht, aber arbeiten da nicht ...๐Ÿ˜‚๐Ÿ˜€[๐Ÿ˜€]
6705Ist von heute, aber halt 3x in englisch๐Ÿ˜‡๐Ÿ™„[๐Ÿ™„]
2058..๐Ÿคฆ๐Ÿ˜…[๐Ÿ˜…]
2072Oder fang endlich mal wieder an, was fรผrs Stud...๐Ÿ˜„๐Ÿ˜ˆ[๐Ÿ˜ˆ]
756Ich hab vom letzten Jahr noch ein Trauma๐Ÿ˜‚๐Ÿ˜†[๐Ÿ˜†]
5966Ist Freitag schon Wochenende?๐Ÿ˜๐Ÿค”[๐Ÿค”]
7007Wenn wir die Klimaanlage im Museum ausschalten...๐Ÿค”๐Ÿคช[๐Ÿคช]
6968Ein Vorbild fรผr uns alle๐Ÿค”๐Ÿคฉ[๐Ÿคฉ]
5665Kein Problem. Du kannst mich eintragen.๐Ÿ™‚๐Ÿ™‚[๐Ÿ™‚]
2641Jonas, du bist gelb markiert! Du kannst keine ...๐Ÿ˜‰๐Ÿ™ƒ[๐Ÿ™ƒ]
1981Da bin ich mir sehr sicher!๐Ÿ˜˜๐Ÿ˜„[๐Ÿ˜„]
3334Das wรคre auf jeden Fall hilfreich! Danke๐Ÿ˜๐Ÿ™‚[๐Ÿ™‚]
4447Ist das Schild denn auch laminiert?!๐Ÿค”๐Ÿ˜[๐Ÿ˜]
288da bin ich mal gespannt wie das aussieht, wen...๐Ÿ˜‚๐Ÿ‘[๐Ÿ‘]
4265Aber es ist doch gebrochen...๐Ÿค”๐Ÿ˜ข[๐Ÿ˜ข]
7405Partey๐Ÿคฆ๐Ÿคช[๐Ÿคช]
...............
4910Vllt gilt das Schild auch nur fรผr schwarze Hun...๐Ÿ˜‚๐Ÿ˜ฑ[๐Ÿ˜ฑ]
2047Ich erprobe mich morgen mit Freya an Addieren ...๐Ÿ™‚๐Ÿ˜„[๐Ÿ˜„]
2650Ich kann leider nicht๐Ÿ˜•๐Ÿ˜•[๐Ÿ˜•]
299Das schafft meine Handykamera aber nicht๐Ÿ˜‚๐Ÿ˜‰[๐Ÿ˜‰]
2499also ihr lieben ich arbeite gerade im arithmeu...๐Ÿ˜๐Ÿ˜…[๐Ÿ˜…]
4783danke๐Ÿ˜Š๐Ÿ’ธ[๐Ÿ’ธ, ๐Ÿ’ธ, ๐Ÿ’ธ]
4040Herrlich๐Ÿคฆ๐Ÿ˜…[๐Ÿค—, ๐Ÿ˜…]
7553babbage, meine Autokerrektur โ€โ™‚๐Ÿค”๐Ÿคฆ[๐Ÿคฆ]
6232Bin erst um 9.15Uhr da.....๐Ÿ‘๐Ÿ”ช[๐Ÿ’ฃ, ๐Ÿ”ช]
3514Dankeschรถn! Dann trage ich dich ein๐Ÿ˜˜๐Ÿ˜Š[๐Ÿ˜Š]
3422Schnitzel๐Ÿ˜…๐Ÿ˜[๐Ÿ˜]
2294Club discreto?๐Ÿ˜‚๐Ÿ˜ณ[๐Ÿ˜ณ, ๐Ÿ˜ณ, ๐Ÿ˜ณ]
8218auch da bin ich etwas neidisch๐Ÿ™„๐Ÿ˜…[๐Ÿ˜…]
4463Mรถchte jemand was vom Cafรฉ Roller?๐Ÿคง๐Ÿ˜‚[๐Ÿ˜‚]
8032Frohe Ostern an Euch alle...๐Ÿ˜Š๐Ÿฅ[๐Ÿฐ, ๐Ÿฃ, ๐Ÿฅ]
4505okay, ihr geht eigentlich ganz verantwortungsv...๐Ÿ˜Š๐Ÿ’[๐Ÿ’]
4810Meine Meinung: Wenn das Personal da ist, dann ...๐Ÿ˜‚๐Ÿคท[๐Ÿคท]
5762Ein shk ist aber billiger๐Ÿ‘ต๐Ÿ™„[๐Ÿ™„]
4247Alle die morgen auf der Arbeit sind: nachher k...๐Ÿ™ˆ๐Ÿ˜‚[๐Ÿ˜‚]
2501ich glaube das wรคre eine stilvolle und romanti...๐Ÿ˜๐Ÿ˜Š[๐Ÿ˜Š]
5278Ich kann das machen๐Ÿ’โœ‹[โœ‹]
5747In der Kรผche im Kรผhlschrank steht eine grรผne T...๐Ÿ˜‰๐Ÿ™‚[๐Ÿ˜‰, ๐Ÿ™‚]
7022Dann halt nicht โ€โ™‚๐Ÿคท๐Ÿคท[๐Ÿคท]
1736Ich hab absolut keine Ahnung davon๐Ÿ˜‚๐Ÿ˜ž[๐Ÿ˜ž]
5909Wir bekommen es bestimmt zu zweit hin, aber we...๐Ÿ™ˆ๐Ÿ˜‰[๐Ÿ˜‰]
5890Pass bloรŸ auf, wenn du die isst mutierst du be...๐Ÿ˜„๐Ÿ˜ฑ[๐Ÿ˜ฑ]
566Das sind 5 MaรŸ pro Abend๐Ÿ™„๐Ÿ˜‹[๐Ÿ˜‹]
6044Sind sehr lecker๐Ÿ˜€๐Ÿ‘Œ[๐Ÿ‘Œ]
1472Ich werde die alle drillen das sage ich dir, a...๐Ÿค”๐Ÿ˜‡[๐Ÿ˜‡]
5702Oh man, das wird ein Tag๐Ÿ˜€๐Ÿ˜„[๐Ÿ˜„]
\n", - "

246 rows ร— 4 columns

\n", - "
" - ], - "text/plain": [ - " message pred trained \\\n", - "8663 mehrere Bรผrostรผhle hintereinander? ๐Ÿ™ˆ ๐Ÿค” \n", - "1862 Ich will auch Essen ๐Ÿ˜‰ ๐Ÿ˜ž \n", - "3921 2x orange ๐Ÿ™‚ ๐Ÿ™‚ \n", - "2222 Man soll auch keine Lernsachen da lassen ๐Ÿ˜€ ๐Ÿ˜€ \n", - "761 Wir kรถnnen auch gerne wieder bei mir zusammen ... ๐Ÿ˜‡ ๐Ÿ˜‰ \n", - "3600 Hab den Bus verpasst, den ich eigentlich nehme... ๐Ÿ˜Ÿ ๐Ÿ‘ธ \n", - "1794 Klaro, gerne ๐Ÿ˜Ž ๐Ÿ™‚ \n", - "861 Sehr gut, Jan! Wenn sich noch jemand melden wรผ... ๐Ÿ˜˜ ๐Ÿ˜‡ \n", - "1469 Jaaa aber mein Mitbewohner ist messy ๐Ÿ˜ฌ ๐Ÿ™Š \n", - "6254 Hallรถchen Es hat ein Herr angerufen, der gern... ๐Ÿ’ ๐Ÿ™‚ \n", - "7968 Iwo ๐Ÿคฆ ๐Ÿ˜‚ \n", - "8102 Keine Ahnung wer da haariger ist ๐Ÿ˜… ๐Ÿง \n", - "6835 Danke ๐Ÿ˜Š ๐Ÿ˜Š \n", - "7135 Wurde bestimmt von Big Bang geklaut ๐Ÿคท ๐Ÿ˜‚ \n", - "4556 Ich kann leider nicht, aber arbeiten da nicht ... ๐Ÿ˜‚ ๐Ÿ˜€ \n", - "6705 Ist von heute, aber halt 3x in englisch ๐Ÿ˜‡ ๐Ÿ™„ \n", - "2058 .. ๐Ÿคฆ ๐Ÿ˜… \n", - "2072 Oder fang endlich mal wieder an, was fรผrs Stud... ๐Ÿ˜„ ๐Ÿ˜ˆ \n", - "756 Ich hab vom letzten Jahr noch ein Trauma ๐Ÿ˜‚ ๐Ÿ˜† \n", - "5966 Ist Freitag schon Wochenende? ๐Ÿ˜ ๐Ÿค” \n", - "7007 Wenn wir die Klimaanlage im Museum ausschalten... ๐Ÿค” ๐Ÿคช \n", - "6968 Ein Vorbild fรผr uns alle ๐Ÿค” ๐Ÿคฉ \n", - "5665 Kein Problem. Du kannst mich eintragen. ๐Ÿ™‚ ๐Ÿ™‚ \n", - "2641 Jonas, du bist gelb markiert! Du kannst keine ... ๐Ÿ˜‰ ๐Ÿ™ƒ \n", - "1981 Da bin ich mir sehr sicher! ๐Ÿ˜˜ ๐Ÿ˜„ \n", - "3334 Das wรคre auf jeden Fall hilfreich! Danke ๐Ÿ˜ ๐Ÿ™‚ \n", - "4447 Ist das Schild denn auch laminiert?! ๐Ÿค” ๐Ÿ˜ \n", - "288 da bin ich mal gespannt wie das aussieht, wen... ๐Ÿ˜‚ ๐Ÿ‘ \n", - "4265 Aber es ist doch gebrochen... ๐Ÿค” ๐Ÿ˜ข \n", - "7405 Partey ๐Ÿคฆ ๐Ÿคช \n", - "... ... ... ... \n", - "4910 Vllt gilt das Schild auch nur fรผr schwarze Hun... ๐Ÿ˜‚ ๐Ÿ˜ฑ \n", - "2047 Ich erprobe mich morgen mit Freya an Addieren ... ๐Ÿ™‚ ๐Ÿ˜„ \n", - "2650 Ich kann leider nicht ๐Ÿ˜• ๐Ÿ˜• \n", - "299 Das schafft meine Handykamera aber nicht ๐Ÿ˜‚ ๐Ÿ˜‰ \n", - "2499 also ihr lieben ich arbeite gerade im arithmeu... ๐Ÿ˜ ๐Ÿ˜… \n", - "4783 danke ๐Ÿ˜Š ๐Ÿ’ธ \n", - "4040 Herrlich ๐Ÿคฆ ๐Ÿ˜… \n", - "7553 babbage, meine Autokerrektur โ€โ™‚ ๐Ÿค” ๐Ÿคฆ \n", - "6232 Bin erst um 9.15Uhr da..... ๐Ÿ‘ ๐Ÿ”ช \n", - "3514 Dankeschรถn! Dann trage ich dich ein ๐Ÿ˜˜ ๐Ÿ˜Š \n", - "3422 Schnitzel ๐Ÿ˜… ๐Ÿ˜ \n", - "2294 Club discreto? ๐Ÿ˜‚ ๐Ÿ˜ณ \n", - "8218 auch da bin ich etwas neidisch ๐Ÿ™„ ๐Ÿ˜… \n", - "4463 Mรถchte jemand was vom Cafรฉ Roller? ๐Ÿคง ๐Ÿ˜‚ \n", - "8032 Frohe Ostern an Euch alle... ๐Ÿ˜Š ๐Ÿฅ \n", - "4505 okay, ihr geht eigentlich ganz verantwortungsv... ๐Ÿ˜Š ๐Ÿ’ \n", - "4810 Meine Meinung: Wenn das Personal da ist, dann ... ๐Ÿ˜‚ ๐Ÿคท \n", - "5762 Ein shk ist aber billiger ๐Ÿ‘ต ๐Ÿ™„ \n", - "4247 Alle die morgen auf der Arbeit sind: nachher k... ๐Ÿ™ˆ ๐Ÿ˜‚ \n", - "2501 ich glaube das wรคre eine stilvolle und romanti... ๐Ÿ˜ ๐Ÿ˜Š \n", - "5278 Ich kann das machen ๐Ÿ’ โœ‹ \n", - "5747 In der Kรผche im Kรผhlschrank steht eine grรผne T... ๐Ÿ˜‰ ๐Ÿ™‚ \n", - "7022 Dann halt nicht โ€โ™‚ ๐Ÿคท ๐Ÿคท \n", - "1736 Ich hab absolut keine Ahnung davon ๐Ÿ˜‚ ๐Ÿ˜ž \n", - "5909 Wir bekommen es bestimmt zu zweit hin, aber we... ๐Ÿ™ˆ ๐Ÿ˜‰ \n", - "5890 Pass bloรŸ auf, wenn du die isst mutierst du be... ๐Ÿ˜„ ๐Ÿ˜ฑ \n", - "566 Das sind 5 MaรŸ pro Abend ๐Ÿ™„ ๐Ÿ˜‹ \n", - "6044 Sind sehr lecker ๐Ÿ˜€ ๐Ÿ‘Œ \n", - "1472 Ich werde die alle drillen das sage ich dir, a... ๐Ÿค” ๐Ÿ˜‡ \n", - "5702 Oh man, das wird ein Tag ๐Ÿ˜€ ๐Ÿ˜„ \n", - "\n", - " emojis \n", - "8663 [๐Ÿค”] \n", - "1862 [๐Ÿ˜ž] \n", - "3921 [๐Ÿ™‚] \n", - "2222 [๐Ÿ˜€] \n", - "761 [๐Ÿ˜‰] \n", - "3600 [๐Ÿ‘ธ] \n", - "1794 [๐Ÿ™‚] \n", - "861 [๐Ÿ˜‡] \n", - "1469 [๐Ÿ˜ฑ, ๐Ÿ˜”, ๐ŸŒš, ๐Ÿ™Š] \n", - "6254 [๐Ÿ™‚] \n", - "7968 [๐Ÿ˜‚] \n", - "8102 [๐Ÿง] \n", - "6835 [๐Ÿ˜Š] \n", - "7135 [๐Ÿ˜‚, ๐Ÿ˜‚, ๐Ÿ˜‚] \n", - "4556 [๐Ÿ˜€] \n", - "6705 [๐Ÿ™„] \n", - "2058 [๐Ÿ˜…] \n", - "2072 [๐Ÿ˜ˆ] \n", - "756 [๐Ÿ˜†] \n", - "5966 [๐Ÿค”] \n", - "7007 [๐Ÿคช] \n", - "6968 [๐Ÿคฉ] \n", - "5665 [๐Ÿ™‚] \n", - "2641 [๐Ÿ™ƒ] \n", - "1981 [๐Ÿ˜„] \n", - "3334 [๐Ÿ™‚] \n", - "4447 [๐Ÿ˜] \n", - "288 [๐Ÿ‘] \n", - "4265 [๐Ÿ˜ข] \n", - "7405 [๐Ÿคช] \n", - "... ... \n", - "4910 [๐Ÿ˜ฑ] \n", - "2047 [๐Ÿ˜„] \n", - "2650 [๐Ÿ˜•] \n", - "299 [๐Ÿ˜‰] \n", - "2499 [๐Ÿ˜…] \n", - "4783 [๐Ÿ’ธ, ๐Ÿ’ธ, ๐Ÿ’ธ] \n", - "4040 [๐Ÿค—, ๐Ÿ˜…] \n", - "7553 [๐Ÿคฆ] \n", - "6232 [๐Ÿ’ฃ, ๐Ÿ”ช] \n", - "3514 [๐Ÿ˜Š] \n", - "3422 [๐Ÿ˜] \n", - "2294 [๐Ÿ˜ณ, ๐Ÿ˜ณ, ๐Ÿ˜ณ] \n", - "8218 [๐Ÿ˜…] \n", - "4463 [๐Ÿ˜‚] \n", - "8032 [๐Ÿฐ, ๐Ÿฃ, ๐Ÿฅ] \n", - "4505 [๐Ÿ’] \n", - "4810 [๐Ÿคท] \n", - "5762 [๐Ÿ™„] \n", - "4247 [๐Ÿ˜‚] \n", - "2501 [๐Ÿ˜Š] \n", - "5278 [โœ‹] \n", - "5747 [๐Ÿ˜‰, ๐Ÿ™‚] \n", - "7022 [๐Ÿคท] \n", - "1736 [๐Ÿ˜ž] \n", - "5909 [๐Ÿ˜‰] \n", - "5890 [๐Ÿ˜ฑ] \n", - "566 [๐Ÿ˜‹] \n", - "6044 [๐Ÿ‘Œ] \n", - "1472 [๐Ÿ˜‡] \n", - "5702 [๐Ÿ˜„] \n", - "\n", - "[246 rows x 4 columns]" - ] - }, - "execution_count": 21, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "testlist = pd.DataFrame({'message': Xt1, 'pred': pred, 'trained': yt1})\n", - "testlist = pd.merge(testlist, emoji_messages['emojis'].to_frame(), left_index=True, right_index=True)\n", - "testlist.to_csv('export.csv')\n", - "testlist" - ] - }, - { - "cell_type": "code", - "execution_count": 22, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "'๐Ÿ™„'" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "display(clf_a.predict(vectorizer.transform([\"Boah Caner\"]))[0])" - ] - }, { "cell_type": "code", "execution_count": 23, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "แผฟ\n" - ] - } - ], + "outputs": [], "source": [ - "print(chr(0x1F3F))" + "pred = clf_a.predict(vectorizer.transform(Xt1))" ] }, { "cell_type": "code", - "execution_count": 24, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "<1x3931 sparse matrix of type ''\n", - "\twith 2 stored elements in Compressed Sparse Row format>" - ] - }, - "execution_count": 24, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ - "vec_train[0]" + "testlist = pd.DataFrame({'message': Xt1, 'pred': mlb.inverse_transform(pred), 'teacher': mlb.inverse_transform(yt1)})\n", + "testlist.to_csv('export.csv')\n", + "testlist" ] }, { "cell_type": "code", - "execution_count": 25, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "application/vnd.jupyter.widget-view+json": { - "model_id": "0b212fa64d5e457d89ad798983002a1c", - "version_major": 2, - "version_minor": 0 - }, - "text/html": [ - "

Failed to display Jupyter Widget of type Text.

\n", - "

\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 Jupyter\n", - " Widgets Documentation for setup instructions.\n", - "

\n", - "

\n", - " If you're reading this message in another frontend (for example, a static\n", - " rendering on GitHub or NBViewer),\n", - " it may mean that your frontend doesn't currently support widgets.\n", - "

\n" - ], - "text/plain": [ - "Text(value='')" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "application/vnd.jupyter.widget-view+json": { - "model_id": "f1d5e13f70a24e80acc95372e3f3db0d", - "version_major": 2, - "version_minor": 0 - }, - "text/html": [ - "

Failed to display Jupyter Widget of type VBox.

\n", - "

\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 Jupyter\n", - " Widgets Documentation for setup instructions.\n", - "

\n", - "

\n", - " If you're reading this message in another frontend (for example, a static\n", - " rendering on GitHub or NBViewer),\n", - " it may mean that your frontend doesn't currently support widgets.\n", - "

\n" - ], - "text/plain": [ - "VBox(children=(Button(description='get smiley', icon='check', style=ButtonStyle(), tooltip='Click me'), Output()))" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], + "outputs": [], "source": [ - "import ipywidgets as widgets \n", "\n", "out = widgets.Output()\n", "\n", @@ -2695,7 +463,7 @@ " with out:\n", " clear_output()\n", " with out:\n", - " display(Markdown(\"# \" + str(clf_a.predict(vectorizer.transform([t.value]))[0])))\n", + " display(Markdown(\"# \" + str(mlb.inverse_transform(clf_a.predict(vectorizer.transform([t.value])))[0])))\n", "\n", "b.on_click(handle_submit)\n", " \n", diff --git a/Project/Tools/whatsapp2csv.sh b/Project/Tools/whatsapp2csv.sh index ad59d5d..178f170 100755 --- a/Project/Tools/whatsapp2csv.sh +++ b/Project/Tools/whatsapp2csv.sh @@ -77,7 +77,7 @@ perform_and_exit echo -e "date\ttime\tperson\tmessage" > "$OUTPUT" # 3. replace separators of whatsapp's txt-format with '\t' as separator for csv # 4. delete all double-quotes. Because people are too stupid to type them in pairs and that breaks some csv-interpreters ๐Ÿคฆโ€โ™‚ -perform_and_exit sed ':a;N;/\n[0-9]\{1,\}\+\//!s/\n/ /;ta;P;D' "$INPUT" | grep -v -E 'left|added|created|end-to-end|Media omitted' | sed 's/, /\t/1; s/- /\t/1; s/: /\t/1;' | tr -d '"' >> "$OUTPUT" +perform_and_exit sed ':a;N;/\n[0-9]\{1,\}\+\//!s/\n/ /;ta;P;D' "$INPUT" | grep -v -E 'left|added|created|end-to-end|Media omitted|import java|> "$OUTPUT" confirm_action