1 Import PANDAS AS PD 2 Import numpy AS NP . 3 . 4 . 5 # normalized magnitude ---- removal effect . 6 . 7 # 3 ways 8 # (1) from the normalized difference 9 # data linearly changes, mapping data to the range [0,1], 10 # X = (X - min) / (max - min) . 11 # too large or too small outlier will influence the outcome 12 # susceptible to the influence of outliers 13 is DEF max_min_sca (Data ): 14 "" " 15 by means of the difference from the normalized data normalized to 16 : param data: original data . 17 : return: data after normalization 18 " "" 19 = Data (Data - data.min ()) / (data.max () - data.min ()) 20 is 21 is return Data 22 is 23 is 24 # (2) normalized standard deviation 25 # by mean and standard deviation data conversion 26 is # X = (X-Mean) / STD 27 DEF stand_sca (data): 28 "" " 29 standard differential standardization 30 : param data: original data 31 is : return: the standard deviation data after 32 " "" 33 is data = ( Data - data.mean ()) / data.std () 34 is 35 return Data 36 37 [ 38 is #[10, 20] 10000 10000 ---- not affect the mean, standard deviation little effect 39 # is not susceptible to the influence of outliers 40 41 is 42 is # (3) fractional scaling standardized 43 # by moving the number of decimal places data conversion between [1,1] --- constant distribution data 44 is # X = X / K 10 ^ 45 # K -----> rounded up (log10 (| x | .max ( ) )) 46 is DEF desc_sca (data): 47 "" " 48 decimal scaling the normalized data 49 : param data: original data 50 : return: data after normalization 51 is " "" 52 is data = data / (10 ** int (NP .ceil (np.log10 (data.abs (). max ())))) 53 is return Data 54 is 55 56 is 57 is # verification: 58 detail pd.read_excel = ( " ./meal_order_detail.xlsx " ) 59 60 Print ( " column index of detail: \ n- " , detail.columns) 61 is # Print ( "Shape of detail: \ n" , detail.shape) 62 is Print ( " before unnormalized: \ n- " , detail.loc [:, " Amounts " ]) 63 is Print ( " maximum and minimum values: \ n- " , detail.loc [:, " Amounts " ] .max (), detail.loc [:, "amounts" ] .Min ()) 64 Print ( " after normalizing \ n- " , max_min_sca (detail.loc [:, " Amounts " ])) 65 Print ( " after normalizing \ n- " , stand_sca (detail.loc [:, " Amounts " ])) 66 Print ( " after normalizing \ n- " , desc_sca (detail.loc [:, " Amounts " ]))
[] Data Analysis Data Mining & standardized way three data - normalized deviation, standard deviation normalized fractional scaling & Standardization
Guess you like
Origin www.cnblogs.com/Tree0108/p/12116093.html
Ranking