{"id":3651,"date":"2023-07-16T08:13:17","date_gmt":"2023-07-16T08:13:17","guid":{"rendered":"https:\/\/statorials.org\/pt\/pandas-combinando\/"},"modified":"2023-07-16T08:13:17","modified_gmt":"2023-07-16T08:13:17","slug":"pandas-combinando","status":"publish","type":"post","link":"https:\/\/statorials.org\/pt\/pandas-combinando\/","title":{"rendered":"Como usar corrwith() no pandas (com exemplos)"},"content":{"rendered":"

<\/p>\n

\n

Voc\u00ea pode usar a fun\u00e7\u00e3o corrwith()<\/strong> em pandas para calcular a correla\u00e7\u00e3o de pares entre colunas num\u00e9ricas com o mesmo nome<\/em> em dois DataFrames de pandas diferentes.<\/span><\/p>\n

Esta fun\u00e7\u00e3o usa a seguinte sintaxe b\u00e1sica:<\/span><\/p>\n

 df1. corrwith<\/span> (df2)<\/span><\/span>\n<\/strong><\/pre>\n
 Nota<\/strong> : Esta fun\u00e7\u00e3o \u00e9 diferente da fun\u00e7\u00e3o corr()<\/strong> , que calcula a correla\u00e7\u00e3o entre duas colunas num\u00e9ricas dentro do mesmo DataFrame.<\/span><\/p>\n
 O exemplo a seguir mostra como usar a fun\u00e7\u00e3o corrwith()<\/strong> na pr\u00e1tica.<\/span><\/p>\n
 Exemplo: como usar corrwith() no Pandas<\/strong><\/span><\/h2>\n
 Vamos supor que temos os dois DataFrames panda a seguir:<\/span><\/p>\n
 import<\/span> pandas as<\/span> pd\n\n#create first DataFrame\n<\/span>df1 = pd. DataFrame<\/span> ({' team<\/span> ': ['A', 'B', 'C', 'D', 'E', 'F'],\n                    ' points<\/span> ': [18, 22, 29, 25, 14, 11],\n                    ' assists<\/span> ': [4, 5, 5, 4, 8, 12],\n                    ' rebounds<\/span> ': [10, 6, 4, 6, 3, 5]})\n\nprint<\/span> (df1)\n\n  team points assists rebounds\n0 to 18 4 10\n1 B 22 5 6\n2 C 29 5 4\n3 D 25 4 6\n4 E 14 8 3\n5 F 11 12 5\n\n#create second DataFrame<\/span>\ndf2 = pd. DataFrame<\/span> ({' team<\/span> ': ['A', 'B', 'C', 'D', 'E', 'F'],\n                    ' points<\/span> ': [22, 25, 27, 35, 25, 20],\n                    ' assists<\/span> ': [15, 13, 8, 8, 5, 8],\n                    ' rebs<\/span> ': [4, 11, 12, 8, 7, 10]})\n\nprint<\/span> (df2)\n\n  team points assists rebs\n0 A 22 15 4\n1 B 25 13 11\n2 C 27 8 12\n3 D 35 8 8\n4 E 25 5 7\n5 F 20 8 10<\/b><\/pre>\n
 Podemos usar a fun\u00e7\u00e3o corrwith()<\/strong> para calcular a correla\u00e7\u00e3o entre colunas num\u00e9ricas com os mesmos nomes nos dois DataFrames:<\/span><\/p>\n
 #calculate correlation between numeric columns with same names in each DataFrame\n<\/span>df1. corrwith<\/span> (df2)\n\npoints 0.677051\nassists -0.478184\nNaN rebounds\nrebs NaN\ndtype:float64<\/span><\/span><\/strong><\/pre>\n
 Pelo resultado podemos ver:<\/span><\/p>\n