名称

SimpleR::Reshape 数据处理转换

说明

接口山寨自R语言的read.table/write.table/merge

还有reshape2包：http://cran.r-project.org/package=reshape2

函数

read_table

支持 从文件或arrayref 按行读入数据，转换后输出新的 文件或arrayref

    my $r = read_table(
        'reshape_src.csv', 
        skip_head=>1, 
        conv_sub => sub { [ "$_[0][0] $_[0][1]", $_[0][2], $_[0][3] ] }, 

        write_filename => '01.read_table.csv', 
        #skip_sub => sub { $_[0][3]<200 }, 
        #return_arrayref => 1, 
        #sep=>',', 
        #charset=>'utf8', 
    );

write_table

将指定数据写入文本文件

    my $d = [ [qw/a b 1/], [qw/c d 2/] ]; 
    write_table($d, 
        file=> 'write_table.csv', 
        header => [ 'ka', 'kb', 'cnt'], 
        #sep => ',', 
        #charset => 'utf8', 
    );

melt

数据调整，参考R语言的reshape2包

    my $r = melt('reshape_src.csv',
        skip_head => 1, 

        names => [ qw/day hour state cnt rank/ ], 

        #skip_sub => sub { $_[0][3]<1000 }, 
        id => [ 0, 1, 2 ],
        measure => [3, 4], 
        melt_filename => '02.melt.csv',
    );

cast

数据重组，参考R语言的reshape2包

    my $r = cast('02.melt.csv', 
        cast_filename => '03.cast.csv', 

        #key 有 cnt / rank 两种
        names => [ qw/day hour state key value/ ], 
        id => [ 0, 1, 2 ],
        measure => 3, 
        value => 4, 

        stat_sub => sub { my ($vlist) = @_; my @temp = sort { $b <=> $a } @$vlist; return $temp[0] }, 

        result_names => [ qw/day hour state cnt rank/ ], 

        #reduce_sub => sub { 
        #   my ($r) = @_;
        #   my $s=0 ; $s+= $_ for @$r; 
        #   return [ $s ];
        #   }, 
    );

注意：

reduce_sub 是在读取数据的过程中处理value，默认是直接push到value列表里

stat_sub 是在数据读取完毕后，对value列表进行最终统计处理

merge

合并两个dataframe，在perl中是二层数组

    my $r = merge( 
        [ [qw/a b 1/], [qw/c d 2/] ], 
        [ [qw/a b 3/], [qw/c d 4/] ], 
        by => [ 0, 1], 
        value => [2], 
    );
    # $r = [["a", "b", 1, 3], ["c", "d", 2, 4]]