Postgresql删除数据库表中重复数据的几种方法详解

DROP TABLE IF EXISTS "public"."devicedata";
CREATE TABLE "public"."devicedata" (
  "Id" varchar(200) COLLATE "pg_catalog"."default" NOT NULL,
  "DeviceId" varchar(200) COLLATE "pg_catalog"."default",
  "Timestamp" int8,
  "DataArray" float4[]
)

CREATE INDEX "timeIndex" ON "public"."devicedata" USING btree (
  "Timestamp" "pg_catalog"."int8_ops" DESC NULLS LAST,
  "DeviceId" COLLATE "pg_catalog"."default" "pg_catalog"."text_ops" ASC NULLS LAST
);

ALTER TABLE "public"."devicedata" ADD CONSTRaiNT "devicedata_pkey" PRIMARY KEY ("Id");

DELETE FROM "DeviceData"
WHERE "Id"
NOT IN (
SELECT max("Id")
FROM "DeviceData_temp"
GROUP BY "DeviceId", "Timestamp"
);

SELECT DISTINCT "Timestamp", "DeviceId"
INTO "DeviceData_temp"
FROM "DeviceData";
-- 删除原表
DROP TABLE "DeviceData";
-- 将新表重命名
ALTER TABLE "DeviceData_temp" RENAME TO "DeviceData";

INSERT INTO "DeviceData_temp"
SELECT * FROM "DeviceData"
on conflict("DeviceId", "Timestamp") DO NOTHING;
-- 删除原表
DROP TABLE "DeviceData";
-- 将新表重命名
ALTER TABLE "DeviceData_temp" RENAME TO "DeviceData";