Transpose columns in postgresql - lateral join?

Question

I'm trying to transpose a few columns in postgresql.

From this data format:

deal id | deal name | create date | lead_date_pipeline_ | warm_lead_date_pipeline_ | d.meeting_1_date_pipeline_ | d.final_meeting_date_pipeline_ | d.contract_sent_date_pipeline_ | d.closed_won_date_pipeline_ etc..

I want to transpose all columns apart from deal name and deal id and get 2 new columns (pipeline stage with all the stages: lead, warm lead, meeting 1, final meeting etc; and date) like this:

deal id | deal name | pipeline stage | data

I was able to do so with the following code:

SELECT d."deal id", d."deal name", v.*
FROM   gg_deals d
     , LATERAL (
   VALUES
      ('create date', d."create date")
    , ('lead_date_pipeline_', d.lead_date_pipeline_)
    , ('warm_lead_date_pipeline_', d.warm_lead_date_pipeline_)
    , ('meeting_1_date_pipeline_', d.meeting_1_date_pipeline_)
    , ('final_meeting_date_pipeline_', d.final_meeting_date_pipeline_)
    , ('contract_sent_date_pipeline_', d.contract_sent_date_pipeline_)
    , ('closed_won_date_pipeline_', d.closed_won_date_pipeline_)
    , ('closed_lost_date_pipeline_', d.closed_lost_date_pipeline_)
   ) v (Pipeline_stage, Date)

However, as we moved the data to redshift, this query no longer works and gives me this error

SQL Error [500310] [42883]: [Amazon](500310) Invalid operation: function values("unknown", timestamp without time zone) does not exist;

how can I get around this? I don't understand where the problem is

Answer 1

A typical work around uses union all :

select deal_id, deal_name, 'create date' as pipeline_stage "create date" as date from gg_deals
union all select deal_id, deal_name, 'lead_date_pipeline_', d.lead_date_pipeline_ from gg_deals
union all select deal_id, deal_name, 'warm_lead_date_pipeline_', d.warm_lead_date_pipeline_ from gg_deals
union all select deal_id, deal_name, 'meeting_1_date_pipeline_', d.meeting_1_date_pipeline_ from gg_deals
...