如何在csv文件的列中将最小/最大值转换为JSON格式（如Python中的格式）？

我想输入几个csv文件，每个文件看起来像这样：

                column1     column2        column3       
18/05/16 11:19  143501      148686          40644           
18/05/16 11:34  144070      144710          79585           
18/05/16 11:49  134070      144713          79588           
18/05/16 12:03  144070      144716          80591           
18/05/16 12:13  144070      144733          79608           
18/05/16 12:23  154790      144733          79608           
18/05/16 12:19  144070      125753          79628           
18/05/16 12:34  144070      144757          79632           

我想获取每个csv文件中每个列的最大值和最小值（不包括日期/时间列），并将数据放入类似以下的结构中：

{
    'csvfile1' :[{

        "column1": [{
            "max": "154790"
            "min": "134070"
            }],
        "column2": [{
            "max": "148686"
            "min": "125753"
            }],
        "column3": [{
            "max": "80591"
            "min": "40644"
            }]
        }],
    'csvfile2' :[{
        <same type of information here>
    }],
    'csvfile3' :[{
        <same type of information here>
    }]
}

到目前为止，这是我的脚本：

max_min_data={}
for name in csvfile_names:
    f = open(name+'.csv', "r")

    #Only want to check columns 1 to 3 not the date/time
    for number in range(1,3):
        f.seek(0)
        column_names = next(f) next(f) # Skip header
        max_value = max(int(row[number]) for row in csv.reader(f))
        min_value = min(int(row[number]) for row in csv.reader(f))

        max_min_data[name]=[column_names[number]]
        max_min_data[name][column_names[number]]=['max','min']
        max_min_data[name][column_names[number]]['max']=max_value
        max_min_data[name][column_names[number]]['min']=min_value

        print "MAX:"
        print '\t\t'+str(max_value)
        print "MIN:"
        print '\t\t'+str(min_value)