weka和meka的.bat批量处理命令小结

布满荆棘的人生 2022-03-16 15:40 83阅读 0赞

1.weka和meka中都可以先导入数据，然后直接选择模型配置各项参数，然后右键复制模型配置参数，粘贴到记事本中再做少量添加就变成了可执行的命令行命令。命令行执行可以输出各个样本预测的概率和标签。  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2pwX3pob3UyNTY_size_16_color_FFFFFF_t_70]2.最笨的命令行方式(需要批量处理的文件个数少时)  
weka和meka均适用。  
2.1.weka

cd D:/Programs/weka-3.9/Weka-3-9
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_40_pinjie.arff -x 10 > D:/chem6_40_10_pinjie.txt
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_50_pinjie.arff -x 10 > D:/chem6_50_10_pinjie.txt
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_60_pinjie.arff -x 10 > D:/chem6_60_10_pinjie_pinjie.txt
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_70_pinjie.arff -x 10 > D:/chem6_70_10_pinjie_pinjie.txt
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_80_pinjie.arff -x 10 > D:/chem6_80_10_pinjie_pinjie.txt
    java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_90_pinjie.arff -x 10 > D:/chem6_90_10_pinjie.txt

2.2.meka

cd D:/Meka1.9.2
    java -cp "./lib/*" meka.classifiers.multilabel.RAkELd -M 1 -k 14 -P 0 -N 0 -S 0 -x 10 -R -verbosity 7 -t data/chem6_700_pinjie.arff -W weka.classifiers.functions.SMO -- -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4" > D:/360Downloads/suss_s2_k1.txt
    java -cp "./lib/*" meka.classifiers.multilabel.RAkELd -M 2 -k 14 -P 0 -N 0 -S 0 -x 10 -R -verbosity 7 -t data/chem6_700_pinjie.arff -W weka.classifiers.functions.SMO -- -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4" > D:/360Downloads/suss_s2_k2.txt
    java -cp "./lib/*" meka.classifiers.multilabel.RAkELd -M 3 -k 14 -P 0 -N 0 -S 0 -x 10 -R -verbosity 7 -t data/chem6_700_pinjie.arff -W weka.classifiers.functions.SMO -- -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4" > D:/360Downloads/suss_s2_k3.txt

3.找规律通过.bat的循环来批量处理文件  
3.1.weka

@echo off
    cd D:/Programs/weka-3.9/Weka-3-9
    for  %%I in (40 50 60 70 80 90) do java weka.classifiers.functions.SMO -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  -t  data/chem6_%%I_pinjie.arff -x 10 > D:/chem6_%%I_10_pinjie.txt
    pause

3.2.meka

@echo off
    cd D:/Meka1.9.2
    for  %%I in (40 50 60 70 80 90) do  java -cp "./lib/*" meka.classifiers.multilabel.RAkELd -M 1 -k 14 -P 0 -N 0 -S 0 -x 10 -R -verbosity 7 -t data/chem6_%%I_pinjie.arff -W weka.classifiers.functions.SMO -- -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" -calibrator "weka.classifiers.functions.Logistic -R 1.0E-8 -M -1 -num-decimal-places 4"  > D:/360Downloads/suss_%%I_s2_k1.txt
    pause

3.3.在.bat文件中使用三重循环

F:
    cd F:/meka-release-1.9.0
    for %%m in (8 10 12 14 16 18 20 25 30) do (
        for %%k in (1 2 3 4 5 6) do  (
            for %%I in (50 60) do java -cp "./lib/*" meka.classifiers.multilabel.RAkEL -M %%m -k %%k -P 0 -N 0 -S 0 -x 10 -t data/%%I.arff -W weka.classifiers.functions.SMO -- -C 1.0 -L 0.001 -P 1.0E-12 -N 0 -V -1 -W 1 -K "weka.classifiers.functions.supportVector.PolyKernel -E 1.0 -C 250007" > E:/Multi_label2019/%%I_%%m_%%k.txt
        )
    )
    pause

3.4.解析SMO(SVM)结果文件,构造DataFrame存档

# -*- coding: utf-8 -*-
    import os
    import pandas as pd
    dir1 = 'C:/Users/Administrator/Desktop/zhouzhou11/'
    result=[]
    for root,dirname,filenames in os.walk(dir1): 
        print(root)
        for filename in filenames:
            dim,m,k=filename.strip().split('.')[0].split('_')
            print(filename)
            with open(root+filename,'r+') as fp:
                content=fp.readlines()
                #str(content[15])
                str(content[18])
                acc=content[15].strip().split('                       ')
                extact_match=content[18].strip().split('                    ')
                result.append([dim,m,k,filename,acc[1],extact_match[1]])
    result_df=pd.DataFrame(result,columns=['维度','M','K','filename','accuracy','extact match'])            
    result_df.to_csv('自己保存的路径!!!',index=False,header=True)

效果  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2pwX3pob3UyNTY_size_16_color_FFFFFF_t_70 1]

# -*- coding: utf-8 -*-
    #统计Excel表格的列数
    str_dict={'A':1,'B':2,'C':3,'D':4,'E':5,'F':6,'G':7,'H':8,'I':9,
     'J':10,'K':11,'L':12,'M':13,'N':14,'O':15,'P':16,'Q':17,
     'R':18,'S':19,'T':20,'U':21,'V':22,'W':23,'X':24,'Y':25,
     'Z':26}
    def input_char(string):
        temp=list(string)
        sum_jishu=0
        base1=[26**2,26,1]
        for i in range(len(temp)):
            sum_jishu+=str_dict[temp[i]]*base1[i]
        print(sum_jishu)
    #input_char('CRG') #2503  
    #input_char('EBC') #3435    
    #input_char('EZP') #4072
    #input_char('HKW') #5717
    #input_char('SXL') #13480
    #input_char('HGZ') #5616
    input_char('AMK') #1024

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2pwX3pob3UyNTY_size_16_color_FFFFFF_t_70]: /images/20220316/d6dd97717f68453abfc0131ee9749313.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2pwX3pob3UyNTY_size_16_color_FFFFFF_t_70 1]: /images/20220316/18ce6e3638d14ce1b327e9a7a553fe91.png