IC 10 optimization

[yixinglu]

What type of PR is this?

• bug
• feature
• enhancement

What problem(s) does this PR solve?

Issue(s) number:

Description:

(root@nebula) [sf100]> MATCH (person:Person)-[:KNOWS]-(friend) WHERE id(person) == "p-32985349068465" WITH person, collect(id(friend)) AS friends MATCH (person:Person)-[:KNOWS*2]-(friend)-[:IS_LOCATED_IN]
->(city:Place) WHERE id(friend) != "p-32985349068465" AND id(friend) NOT IN friends WITH person, city, friend, datetime(friend.Person.birthday) as birthday WHERE (birthday.month==3 AND birthday.day>=21) O
R         (birthday.month==(3%12)+1 AND birthday.day<22) WITH DISTINCT friend AS f, city AS c, person AS p OPTIONAL MATCH (f)<-[:POST_HAS_CREATOR]-(post:Post) OPTIONAL MATCH (p)-[:HAS_INTEREST]->(m), (pos
t:Post)-[:HAS_TAG]->(m) with f AS friend, c AS city, post, case when count(m) > 0 then 1 else 0 END as cnt with friend, city, sum(cnt) AS commonPostCount, count(post) AS postCount RETURN toInteger(substr(
id(friend), 2)) AS personId,        friend.Person.firstName AS personFirstName,        friend.Person.lastName AS personLastName,        commonPostCount - (postCount - commonPostCount) AS commonInterestSco
re,        friend.Person.gender AS personGender,        city.Place.name AS personCityName ORDER BY commonInterestScore DESC, personId ASC LIMIT 10
+----------------+-----------------+----------------+---------------------+--------------+----------------+
| personId       | personFirstName | personLastName | commonInterestScore | personGender | personCityName |
+----------------+-----------------+----------------+---------------------+--------------+----------------+
| 447564         | "Monja"         | "Simon"        | 144                 | "female"     | "Antananarivo" |
| 15393163189959 | "Amit"          | "Singh"        | 104                 | "male"       | "Bandra"       |
| 28587302727145 | "R."            | "Khan"         | 96                  | "female"     | "Kasaragod"    |
| 15393162886890 | "Nikhil"        | "Singh"        | 87                  | "female"     | "Kerala"       |
| 21990232659515 | "Rahul"         | "Kumar"        | 86                  | "female"     | "Tirunelveli"  |
| 13194139647349 | "Sanjay"        | "Khan"         | 80                  | "male"       | "Barasat"      |
| 2199023256816  | "K."            | "Bose"         | 57                  | "female"     | "Pondicherry"  |
| 24189256000791 | "K."            | "Sharma"       | 48                  | "female"     | "Gaya"         |
| 6597069783855  | "Rahul"         | "Singh"        | 41                  | "female"     | "Aligarh"      |
| 13194139771673 | "R."            | "Singh"        | 40                  | "female"     | "Kanpur"       |
+----------------+-----------------+----------------+---------------------+--------------+----------------+
Got 10 rows (time spent 5.587754s/5.589362496s)

Execution Plan (optimize time 1117 us)


Sun, 04 Dec 2022 23:15:44 CST

(root@nebula) [sf100]> """
                    -> MATCH (person:Person)-[:KNOWS]-(friend)
                    -> WHERE id(person) == "p-32985349068465"
                    -> WITH person, collect(id(friend)) AS friends
                    -> MATCH (person:Person)-[:KNOWS*2]-(friend)-[:IS_LOCATED_IN]->(city:Place)
                    -> WHERE id(friend) != "p-32985349068465" AND id(friend) NOT IN friends
                    -> WITH person, city, friend, datetime(friend.Person.birthday) as birthday
                    -> WHERE (birthday.month==3 AND birthday.day>=21) OR
                    ->         (birthday.month==(3%12)+1 AND birthday.day<22)
                    -> WITH DISTINCT friend AS f, city AS c, person AS p
                    -> OPTIONAL MATCH (f)<-[:POST_HAS_CREATOR]-(post:Post)
                    -> OPTIONAL MATCH (p)-[:HAS_INTEREST]->(m)<-[:HAS_TAG]-(post)
                    -> with f AS friend, c AS city, post, case when count(m) > 0 then 1 else 0 END as cnt
                    -> with friend, city, sum(cnt) AS commonPostCount, count(post) AS postCount
                    -> RETURN toInteger(substr(id(friend), 2)) AS personId,
                    ->        friend.Person.firstName AS personFirstName,
                    ->        friend.Person.lastName AS personLastName,
                    ->        commonPostCount - (postCount - commonPostCount) AS commonInterestScore,
                    ->        friend.Person.gender AS personGender,
                    ->        city.Place.name AS personCityName
                    -> ORDER BY commonInterestScore DESC, personId ASC
                    -> LIMIT 10
                    -> """
+----------------+-----------------+----------------+---------------------+--------------+----------------+
| personId       | personFirstName | personLastName | commonInterestScore | personGender | personCityName |
+----------------+-----------------+----------------+---------------------+--------------+----------------+
| 447564         | "Monja"         | "Simon"        | 144                 | "female"     | "Antananarivo" |
| 15393163189959 | "Amit"          | "Singh"        | 104                 | "male"       | "Bandra"       |
| 28587302727145 | "R."            | "Khan"         | 96                  | "female"     | "Kasaragod"    |
| 15393162886890 | "Nikhil"        | "Singh"        | 87                  | "female"     | "Kerala"       |
| 21990232659515 | "Rahul"         | "Kumar"        | 86                  | "female"     | "Tirunelveli"  |
| 13194139647349 | "Sanjay"        | "Khan"         | 80                  | "male"       | "Barasat"      |
| 2199023256816  | "K."            | "Bose"         | 57                  | "female"     | "Pondicherry"  |
| 24189256000791 | "K."            | "Sharma"       | 48                  | "female"     | "Gaya"         |
| 6597069783855  | "Rahul"         | "Singh"        | 41                  | "female"     | "Aligarh"      |
| 13194139771673 | "R."            | "Singh"        | 40                  | "female"     | "Kanpur"       |
+----------------+-----------------+----------------+---------------------+--------------+----------------+
Got 10 rows (time spent 2m34.217948s/2m34.214210283s)

Execution Plan (optimize time 1054 us)


Sun, 04 Dec 2022 23:19:21 CST

How do you solve it?

Special notes for your reviewer, ex. impact of this fix, design document, etc: