Welcome to OStack Knowledge Sharing Community for programmer and developer-Open, Learning and Share
Welcome To Ask or Share your Answers For Others

Categories

0 votes
3.5k views
in Technique[技术] by (71.8m points)

python - Query timeseries from QuestDb using Pandas with parameters

I am trying to load timeseries data from QuestDb to Pandas Dataframe. I'm trying to use Postgres driver like

import pandas as pd
from sqlalchemy import create_engine
from datetime import datetime

engine = create_engine('postgresql://admin:quest@localhost:8812/mydb')
df = pd.read_sql_query(
  "select * from cases where ts between %(dstart)s and %(dfinish)s",
  con=engine, 
  params={"dstart":datetime(2020,12,24,16,0),"dfinish":datetime(2021,1,1,0,0)})

But got back

DatabaseError: (psycopg2.DatabaseError) between/and parameters must be constants
LINE 1: ...etry where ts between '2020-12-24T16:00:00'::timestamp ...

I also tried to use > and < instead of BETWEEN

df = pd.read_sql_query(
  "select * from cases where ts > %(dstart)s and ts < %(dfinish)s",
  con=engine, 
  params={"dstart":datetime(2014,6,24,16,0),"dfinish":datetime(2014,6,24,17,0)})

But still no luck

DatabaseError: (psycopg2.DatabaseError) unsupported class
LINE 1: ... > '2020-06-24T16:00:00'::timestamp and ts < '2021-01-0...

It works without parameters so it's not entirely wrong but I'm missing something here.


与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…
Welcome To Ask or Share your Answers For Others

1 Answer

0 votes
by (71.8m points)

QuestDb has Postgres driver support but is not full SQL query compatible. Few bits don't work so BETWEEN apparently can only be used with constants. Also when you use python datetime parameters they are converted to '2020-06-24T16:00:00'::timestamp in the query which is also not supported by QuestDb

The workaround is to pass string parameters and convert them to timestamp in the query itself, something like

df = pd.read_sql_query(
  "select * from cases where ts > to_timestamp(%(dstart)s, 'yyyy-MM-dd HH:mm:ss') " + 
  "and ts < to_timestamp(%(dfinish)s, 'yyyy-MM-dd HH:mm:ss')",
  con=engine, 
  params={"dstart":datetime(2020,12,24,16,0).strftime("%Y-%m-%d %H:%M:%S"),
  "dfinish":datetime(2021,6,24,17,0).strftime("%Y-%m-%d %H:%M:%S")})

That's the same as in the Trenton mentioned in the comments.


与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…
Welcome to OStack Knowledge Sharing Community for programmer and developer-Open, Learning and Share
Click Here to Ask a Question

...