Hive之count函数的特别用法

create table test_middle_data.spe_count_test(
    name string,
    sex  string,
    is_valid string
)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY ‘,‘;


insert into test_middle_data.spe_count_test values(‘jack01‘,‘man‘,‘N‘);
insert into test_middle_data.spe_count_test values(‘jack02‘,‘woman‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack03‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack04‘,‘woman‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack05‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack06‘,‘woman‘,‘N‘);
insert into test_middle_data.spe_count_test values(‘jack07‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack08‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack09‘,‘man‘,‘N‘);
insert into test_middle_data.spe_count_test values(‘jack10‘,‘woman‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack11‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack12‘,‘man‘,‘Y‘);
insert into test_middle_data.spe_count_test values(‘jack13‘,‘woman‘,‘N‘);

select a.sex,a.is_valid_y,b.total_num
FROM(
    SELECT sex,count(1) is_valid_y
    FROM test_middle_data.spe_count_test
    where is_valid = ‘Y‘
    group by sex
) a
inner join (
    SELECT sex, count(1) total_num
    FROM test_middle_data.spe_count_test
    group by sex
) b 
on a.sex = b.sex

select sex,count(case when is_valid = ‘Y‘ then is_valid else NULL end) is_valid_y,
       count(1)total_num
FROM test_middle_data.spe_count_test
group by sex