Question

我试图创建视图或表格，其中是从点缓冲区生成的多边形，并根据年份对所有多边形进行分组。源表中的数据字段：

point_id(int), point_created(date), geom geometry(Point,3301)
    1, 2014-05-09, point
    2, 2015-01-01, point
    2, 2015-02-05, point
    3, 2016-02-05, point
    4, 2017-02-10, point

我能够创建表格，其中我按年份对所有点进行分组，并将缓冲区生成为多面，但我需要的是根据年份之间的日期对组进行分组（组合一年以上），因此表必须如下所示：

polygon(geom), nr_of_features(int), year(string)
 1, 1, 2014(all point from 2014)   
 2, 3, 2015(all points from 2014 to 2015)
 3, 4, 2016(all points from 2014 to 2016)
 4, 5, 2017(all points from 2014 to 2017)

脚本I，我现在正在使用：

CREATE TABLE my_new_table as 
SELECT ST_Union(ST_Buffer(geom,10))::geometry(MultiPolygon,3301) as polygon,count(point_id)::integer as nr_of_features,
extract(year from point_created) as year
FROM my_table
group by year;

欢迎任何帮助。

Answer 1

如果我理解你，你需要这样的东西：

--just for create few years
WITH RECURSIVE dates AS (
    SELECT '2014-12-31'::timestamp years
    UNION ALL
    SELECT years + interval '1 year' FROM dates WHERE years < '2019-12-31'
)

SELECT
    EXTRACT(YEAR FROM d.years) years,
    p.*
FROM 
    points p
INNER JOIN
    dates d ON p.date <= d.years
;

现在，您可以使用years列来汇总这些点。

Answer 2

由于点日期介于1994年和当前日期之间，我不得不稍微修改这个脚本，因此我将样本值“2014-12-31”改为2013-12-31和“2019-12-31”改为current_date 。它现在似乎有效，谢谢你。我有13000个点，使用此查询创建表需要173毫秒：

WITH RECURSIVE dates AS (
    SELECT '1993-12-31'::timestamp years
    UNION ALL
    SELECT years + interval '1 year' FROM dates WHERE years < current_date
)

SELECT
    EXTRACT(YEAR FROM d.years) years,
    ST_Multi(ST_Collect(geom))::geometry(MultiPoint,3301) as polygon, count(years)::int as nr_features
FROM 
    my_table p
INNER JOIN
    dates d ON p.point_created <= d.years
   group by years
   order by years asc;

使用日期过滤器从点创建缓冲区postgis表

2 个答案: