简体   繁体   English

SQL数据库,数据透视表需要最大日期

[英]SQL database, Need max date for pivot table

Here is my issue. 这是我的问题。 I have a sub query that has several joins (one of the tables do not have a person_id) I have a pivot table created off this data and receive the results of 6 rows for my test data. 我有一个具有几个联接的子查询(其中一个表没有person_id),我有一个根据数据创建的数据透视表,并接收了6行测试数据的结果。 The problem is I need 2 rows.. the two unique person_id's with the most recent lab data. 问题是我需要2行..具有最新实验室数据的两个唯一person_id。 Here is my Query so far: 到目前为止,这是我的查询:

SELECT person_id, first_name,
middle_name, last_name, address_line_1,
address_line_2, city, state, zip,
date_of_birth, sex, enc_id, ClinicalComments,
RecheckMonths, SignedName, [CHOLESTEROL],[HDL CHOLESTEROL],
[LDL CHOLESTEROL,CALCULATED], [TRIGLYCERIDES], coll_date_time
from 
SELECT person_id, first_name,
middle_name, last_name, address_line_1,
address_line_2, city, state, zip,
date_of_birth, sex, enc_id, ClinicalComments,
RecheckMonths, SignedName, [CHOLESTEROL],[HDL CHOLESTEROL],
[LDL CHOLESTEROL,CALCULATED], [TRIGLYCERIDES], coll_date_time
from 
(SELECT p.person_id as person_id,
    p.city as city,
    p.first_name as first_name,
    p.middle_name as middle_name,
    p.last_name as last_name,
    p.address_line_1 as address_line_1,
    p.address_line_2 as address_line_2,
    p.state as state,
    p.zip as zip,
    p.date_of_birth as date_of_birth,
    p.sex as sex,
    cl.enc_id as enc_id,
    cl.ClinicalComments as ClinicalComments,
    cl.RecheckMonths as RecheckMonths,
    cl.SignedName as SignedName,
    lab.result_desc as result_desc,
    lab.observ_value as observ_value,
    lab_results_obr_p.coll_date_time as coll_date_time
            FROM person p  
            LEFT OUTER JOIN card_lipid_ cl ON p.person_id = cl.person_id
            left outer join CANEA_CARD_LIPIDS_ ccl on cl.enc_id=ccl.enc_id 
            left outer join lab_results_obx lab on ccl.person_id = lab.person_id
            left outer join lab_results_obr_p on lab.unique_obr_num = lab_results_obr_p.unique_obr_num
WHERE ccl.LIPIDS='1'
AND lab.delete_ind='N'
AND lab.result_desc in ('CHOLESTEROL','HDL CHOLESTEROL','LDL CHOLESTEROL,CALCULATED','TRIGLYCERIDES')) SourceTable
PIVOT
(max(observ_value) for result_desc in ([CHOLESTEROL],[HDL CHOLESTEROL],[LDL CHOLESTEROL,CALCULATED],[TRIGLYCERIDES])) AS PivotTable

This gives results of 这给出了

ClinicalComments    RecheckMonths   SignedName  CHOLESTEROL HDL CHOLESTEROL LDL CHOLESTEROL;CALCULATED  TRIGLYCERIDES   coll_date_time
None    2   Dr Singer   112 35  52.2    126 10/26/2010 11:08
None    1   Dr Singer   106 41  42  114 1/11/2011 10:41
None    0   Dr Singer   112 3           4/26/2011 12:00
None    1   Dr Singer   96  37  48  58  10/8/2012 9:45
None    1   Dr Singer   103 48  36  109 8/30/2011 12:00
None    1   Dr Singer   102 41  37  120 2/14/2012 11:20

I want it to return lines 4 and 6 only? 我只希望它返回第4行和第6行吗? any ideas? 有任何想法吗?

There are a few ways that I can see that you return the rows with the most recent date. 有几种方法可以让我看到您返回具有最新日期的行。

First way, you can wrap the entire query in a select so you apply a row_number() and partition by the name, ordering by date: 第一种方法,您可以将整个查询包装在一个select中,以便您按名称应用row_number()并按名称进行分区,并按日期排序:

select person_id, first_name,
    middle_name, last_name, address_line_1,
    address_line_2, city, state, zip,
    date_of_birth, sex, enc_id, ClinicalComments,
    RecheckMonths, SignedName, [CHOLESTEROL],[HDL CHOLESTEROL],
    [LDL CHOLESTEROL,CALCULATED], [TRIGLYCERIDES], coll_date_time
from
(
    SELECT person_id, first_name,
        middle_name, last_name, address_line_1,
        address_line_2, city, state, zip,
        date_of_birth, sex, enc_id, ClinicalComments,
        RecheckMonths, SignedName, [CHOLESTEROL],[HDL CHOLESTEROL],
        [LDL CHOLESTEROL,CALCULATED], [TRIGLYCERIDES], coll_date_time,
        ROW_NUMBER() over(partition by person_id order by coll_date_time desc) rn 
    from 
    (
        SELECT p.person_id as person_id,
            p.city as city,
            p.first_name as first_name,
            p.middle_name as middle_name,
            p.last_name as last_name,
            p.address_line_1 as address_line_1,
            p.address_line_2 as address_line_2,
            p.state as state,
            p.zip as zip,
            p.date_of_birth as date_of_birth,
            p.sex as sex,
            cl.enc_id as enc_id,
            cl.ClinicalComments as ClinicalComments,
            cl.RecheckMonths as RecheckMonths,
            cl.SignedName as SignedName,
            lab.result_desc as result_desc,
            lab.observ_value as observ_value,
            lab_results_obr_p.coll_date_time as coll_date_time
        FROM person p  
        LEFT OUTER JOIN card_lipid_ cl 
            ON p.person_id = cl.person_id
        left outer join CANEA_CARD_LIPIDS_ ccl 
            on cl.enc_id=ccl.enc_id 
        left outer join lab_results_obx lab 
            on ccl.person_id = lab.person_id
        left outer join lab_results_obr_p
            on lab.unique_obr_num = lab_results_obr_p.unique_obr_num
        WHERE ccl.LIPIDS='1'
            AND lab.delete_ind='N'
            AND lab.result_desc in ('CHOLESTEROL','HDL CHOLESTEROL','LDL CHOLESTEROL,CALCULATED','TRIGLYCERIDES')
    ) SourceTable
    PIVOT
    (
        max(observ_value) 
        for result_desc in ([CHOLESTEROL],[HDL CHOLESTEROL],[LDL CHOLESTEROL,CALCULATED],[TRIGLYCERIDES])
    ) AS PivotTable
) src
where rn = 1;

Or you could select the max(coll_date_time0 in your subquery, then the most recent value is returned: 或者,您可以在子查询中选择max(coll_date_time0 ,然后返回最新值:

SELECT person_id, first_name,
    middle_name, last_name, address_line_1,
    address_line_2, city, state, zip,
    date_of_birth, sex, enc_id, ClinicalComments,
    RecheckMonths, SignedName, [CHOLESTEROL],[HDL CHOLESTEROL],
    [LDL CHOLESTEROL,CALCULATED], [TRIGLYCERIDES], coll_date_time
from 
(
    SELECT p.person_id as person_id,
        p.city as city,
        p.first_name as first_name,
        p.middle_name as middle_name,
        p.last_name as last_name,
        p.address_line_1 as address_line_1,
        p.address_line_2 as address_line_2,
        p.state as state,
        p.zip as zip,
        p.date_of_birth as date_of_birth,
        p.sex as sex,
        cl.enc_id as enc_id,
        cl.ClinicalComments as ClinicalComments,
        cl.RecheckMonths as RecheckMonths,
        cl.SignedName as SignedName,
        lab.result_desc as result_desc,
        lab.observ_value as observ_value,
        lab_results_obr_p.coll_date_time as coll_date_time
    FROM person p  
    LEFT OUTER JOIN card_lipid_ cl 
        ON p.person_id = cl.person_id
    left outer join CANEA_CARD_LIPIDS_ ccl 
        on cl.enc_id=ccl.enc_id 
    left outer join lab_results_obx lab 
        on ccl.person_id = lab.person_id
    left outer join
    (
        select MAX(coll_date_time) coll_date_time,
            unique_obr_num
        from lab_results_obr_p
        group by unique_obr_num
    ) lab_results_obr_p
        on lab.unique_obr_num = lab_results_obr_p.unique_obr_num
    WHERE ccl.LIPIDS='1'
        AND lab.delete_ind='N'
        AND lab.result_desc in ('CHOLESTEROL','HDL CHOLESTEROL','LDL CHOLESTEROL,CALCULATED','TRIGLYCERIDES')
) SourceTable
PIVOT
(
    max(observ_value) 
    for result_desc in ([CHOLESTEROL],[HDL CHOLESTEROL],[LDL CHOLESTEROL,CALCULATED],[TRIGLYCERIDES])
) AS PivotTable

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM