{"id":3121,"date":"2023-07-19T03:04:16","date_gmt":"2023-07-19T03:04:16","guid":{"rendered":"https:\/\/statorials.org\/tr\/panda-tren-testi\/"},"modified":"2023-07-19T03:04:16","modified_gmt":"2023-07-19T03:04:16","slug":"panda-tren-testi","status":"publish","type":"post","link":"https:\/\/statorials.org\/tr\/panda-tren-testi\/","title":{"rendered":"Pandas dataframe'den e\u011fitim ve test seti nas\u0131l olu\u015fturulur?"},"content":{"rendered":"

<\/p>\n

\n

Makine \u00f6\u011frenimi modellerini veri k\u00fcmelerine yerle\u015ftirirken genellikle veri k\u00fcmesini iki gruba ay\u0131r\u0131r\u0131z:<\/span><\/p>\n

1. E\u011fitim seti:<\/strong> modeli e\u011fitmek i\u00e7in kullan\u0131l\u0131r (orijinal veri setinin %70-80’i)<\/span><\/p>\n

2. Test seti:<\/strong> model performans\u0131na ili\u015fkin tarafs\u0131z bir tahmin elde etmek i\u00e7in kullan\u0131l\u0131r (orijinal veri setinin %20-30’u)<\/span><\/p>\n

Python’da bir pandan\u0131n DataFrame’ini e\u011fitim seti ve test seti olarak ay\u0131rman\u0131n iki yayg\u0131n yolu vard\u0131r:<\/span><\/p>\n

Y\u00f6ntem 1: Sklearn’in train_test_split() i\u015flevini kullan\u0131n<\/strong><\/span><\/p>\n

 from<\/span> sklearn. model_selection<\/span> import<\/span> train_test_split\n\ntrain, test = train_test_split(df, test_size= 0.2<\/span> , random_state= 0<\/span> )<\/strong><\/pre>\n
 Y\u00f6ntem 2: pandalardan sample() \u00f6\u011fesini kullan\u0131n<\/strong><\/span><\/p>\n
 train = df. sample<\/span> (frac= 0.8<\/span> , random_state= 0<\/span> )\ntest = df. drop<\/span> ( train.index<\/span> )<\/strong><\/pre>\n
 A\u015fa\u011f\u0131daki \u00f6rnekler, her y\u00f6ntemin a\u015fa\u011f\u0131daki pandalar DataFrame ile nas\u0131l kullan\u0131laca\u011f\u0131n\u0131 g\u00f6sterir:<\/span><\/p>\n
 import<\/span> pandas as<\/span> pd\nimport<\/span> numpy as<\/span> np\n\n#make this example reproducible\n<\/span>n.p. random<\/span> . seeds<\/span> (1)\n\n#create DataFrame with 1,000 rows and 3 columns\n<\/span>df = pd. DataFrame<\/span> (<\/span> {' x1<\/span> ': np.random.randint<\/span> (30,size=1000),\n                   ' x2<\/span> ': np. random<\/span> . randint<\/span> (12, size=1000),\n                   ' y<\/span> ': np. random<\/span> . randint<\/span> (2, size=1000)})\n\n#view first few rows of DataFrame<\/span>\ndf. head<\/span> ()\n\n        x1 x2 y\n0 5 1 1\n1 11 8 0\n2 12 4 1\n3 8 7 0\n4 9 0 0\n<\/strong><\/pre>\n
 \u00d6rnek 1: sklearn’den train_test_split() i\u015flevini kullan\u0131n<\/strong><\/span><\/h3>\n
 A\u015fa\u011f\u0131daki kod, pandalar\u0131n DataFrame’ini e\u011fitim ve test k\u00fcmelerine b\u00f6lmek i\u00e7in sklearn’in<\/strong> train_test_split()<\/strong> i\u015flevinin nas\u0131l kullan\u0131laca\u011f\u0131n\u0131 g\u00f6sterir:<\/span><\/span><\/p>\n
 from<\/span> sklearn. model_selection<\/span> import<\/span> train_test_split\n\n#split original DataFrame into training and testing sets\n<\/span>train, test = train_test_split(df, test_size= 0.2<\/span> , random_state= 0<\/span> )\n\n#view first few rows of each set<\/span>\nprint<\/span> ( train.head<\/span> ())\n\n     x1 x2 y\n687 16 2 0\n500 18 2 1\n332 4 10 1\n979 2 8 1\n817 11 1 0\n\nprint<\/span> ( test.head<\/span> ())\n\n     x1 x2 y\n993 22 1 1\n859 27 6 0\n298 27 8 1\n553 20 6 0\n672 9 2 1\n\n#print size of each set<\/span>\nprint<\/span> (train. shape<\/span> , test. shape<\/span> )\n\n(800, 3) (200, 3)\n<\/strong><\/pre>\n
 Sonu\u00e7tan iki setin olu\u015fturuldu\u011funu g\u00f6rebiliriz:<\/span><\/p>\n