LPCE
This is for SIGMOD 2023 paper: "Speeding Up End-to-end Query Execution via Learning-based Progressive Cardinality Estimation".
LPCE-I
- In
LPCE-I
directory, please check theREADME
for how to train and test LPCE-I model.
LPCE-R
- In
LPCE-R
directory, please check theREADME
for how to train and test LPCE-R model.
Workload
- In
Workload
directory, please check thejoin-six.sql
andjoin-eight.sql
for experiments.
Distill
- In
Distill
directory, please check theREADME
for how to distill LPCE model.
Use LPCE in Postgres
- In
LPCE_inPostgres
directory, please check how to adopt LPCE in PostgreSQL for query speedup.
Technical report
techreport.pdf
fills in some details not shown in SIGMOD paper due to the limited space.
From author
- Current release is adopting LPCE-I in PostgreSQL. We will release the adoption of LPCE-R in PostgreSQL soon.
- Current release was tested on PostgreSQL 13.0 version.