{"id":2629,"date":"2023-07-21T10:30:26","date_gmt":"2023-07-21T10:30:26","guid":{"rendered":"https:\/\/statorials.org\/pl\/takie-percentyle\/"},"modified":"2023-07-21T10:30:26","modified_gmt":"2023-07-21T10:30:26","slug":"takie-percentyle","status":"publish","type":"post","link":"https:\/\/statorials.org\/pl\/takie-percentyle\/","title":{"rendered":"Jak obliczy\u0107 percentyle w sas: z przyk\u0142adami"},"content":{"rendered":"

<\/p>\n

\n

Oto trzy najcz\u0119stsze sposoby obliczania percentyli dla zbioru danych w SAS-ie:<\/span><\/p>\n

Metoda 1: Oblicz konkretn\u0105 warto\u015b\u0107 percentyla<\/strong><\/span><\/p>\n

 \/*calculate 70th percentile value for var1*\/\n<\/span>proc univariate<\/span> data<\/span> =original_data;\n    <\/span>var<\/span> var1;\n    <\/span>output out<\/span> =percentile_data\n<\/span>pctlpts<\/span> =<\/span> 70<\/span>\n    pctlpre<\/span> = P_;\n<\/span>run<\/span> ;\n<\/span><\/strong><\/span><\/pre>\n
 Metoda 2: Oblicz wiele okre\u015blonych warto\u015bci percentylowych<\/strong><\/span><\/p>\n
 \/*calculate 70th, 80th, and 90th percentile value for var1*\/\n<\/span>proc univariate<\/span> data<\/span> =original_data;\n    var<\/span> var1;\n    output out<\/span> =percentile_data\n    pctlpts<\/span> = 70 80 90<\/span>\n    pctlpre<\/span> = P_;\nrun<\/span> ;<\/strong><\/span><\/pre>\n
 Metoda 3: oblicz percentyle wed\u0142ug grup<\/strong><\/span><\/p>\n
 \/*sort original data by var2*\/\n<\/span>proc sort<\/span> data<\/span> = original_data;\n    by<\/span> var2;\nrun<\/span> ;\n\n\/*calculate percentiles for var1 grouped by var2*\/\n<\/span>proc univariate<\/span> data<\/span> =original_data;\n    var<\/span> var1;\n    by<\/span> var2;\n    output<\/span> out<\/span> =percentile_data\n    pctlpts<\/span> = 70, 80, 90\n    pctlpre<\/span> = P_;\nrun<\/span> ;<\/strong><\/span><\/pre>\n
 Uwaga<\/strong> : Instrukcja pctlpts<\/strong> okre\u015bla percentyle do obliczenia, a instrukcja pctlpre<\/strong> okre\u015bla przedrostek u\u017cywany dla percentyli na wyj\u015bciu.<\/span><\/p>\n
 Poni\u017csze przyk\u0142ady pokazuj\u0105, jak u\u017cywa\u0107 ka\u017cdej metody z nast\u0119puj\u0105cym zbiorem danych w SAS-owym:<\/span> <\/p>\n
 \/*create dataset*\/\ndata<\/span> original_data;\n    input<\/span> team $points;\n    datalines<\/span> ;\nAT 12\nAt 15\nAt 16\nAt 21\nAt 22\nAt 25\nAt 29\nAt 31\nB16\nB22\nB25\nB29\nB 30\nB 31\nB 33\nB 38\n;\nrun<\/span> ;\n\n\/*view dataset*\/\n<\/span>proc print<\/span> data<\/span> = original_data;<\/span><\/span><\/strong> <\/pre>\n
\"\"<\/p>\n
 Przyk\u0142ad 1: Oblicz konkretn\u0105 warto\u015b\u0107 percentyla<\/strong><\/span><\/h3>\n
 Poni\u017cszy kod pokazuje, jak obliczy\u0107 70. percentyl dla zmiennej punkt\u00f3w<\/strong> :<\/span> <\/p>\n
 \/*calculate 70th percentile value for points*\/\n<\/span>proc univariate<\/span> data<\/span> =original_data;\n    var<\/span> points;\n    output out<\/span> =percentile_data\n    pctlpts<\/span> = 70<\/span>\n    pctlpre<\/span> = P_;\nrun<\/span> ;\n\n\/*view results*\/<\/span>\nproc print<\/span> data<\/span> =percentile_data;\n<\/strong><\/span><\/pre>\n
\"\"<\/p>\n
 Warto\u015b\u0107 na 70. percentylu okazuje si\u0119 wynosi\u0107 30<\/strong> .<\/span><\/p>\n
 Przyk\u0142ad 2: Oblicz wiele okre\u015blonych warto\u015bci percentylowych<\/strong><\/span><\/h3>\n
 Poni\u017cszy kod pokazuje, jak obliczy\u0107 warto\u015bci 70., 80. i 90. percentyla dla zmiennej punkt\u00f3w<\/strong> :<\/span> <\/p>\n
 \/*calculate 70th, 80th, and 90th percentile value for points*\/\n<\/span>proc univariate<\/span> data<\/span> =original_data;\n    var<\/span> points;\n    output out<\/span> =percentile_data\n    pctlpts<\/span> = 70 80 90<\/span>\n    pctlpre<\/span> = P_;\nrun<\/span> ;<\/strong><\/span> <\/pre>\n
\"\"<\/p>\n
 Oto jak zinterpretowa\u0107 wynik:<\/span><\/p>\n