[HIVEMALL-274] Wrong target variable name in the step-by-step tutorial

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 0.6.0
• Assignee: Makoto Yui
• Reporter: Aki Ariga
• Created: 31/Oct/19 06:54
• Updated: 31/Oct/19 07:46
• Resolved: 31/Oct/19 07:45

Description

In the step-by-step tutorial, the target column name of regression should be `num_purchases` instead of `label`.

create table if not exists regressor as
select
  train_regressor(
    features, -- feature vector
    num_purchases, -- target value
    '-loss_function squared -optimizer AdaGrad -regularization l2' -- hyper-parameters
  ) as (feature, weight)
from
  training
;

with features_exploded as (
  select
    id,
    extract_feature(fv) as feature,
    extract_weight(fv) as value
  from
    training t1 
    LATERAL VIEW explode(features) t2 as fv
),
predictions as (
  select
    t1.id,
    sum(p1.weight * t1.value) as predicted_num_purchases
  from
    features_exploded t1
    LEFT OUTER JOIN regressor p1 ON (t1.feature = p1.feature)
  group by
    t1.id
)
select
  rmse(t1.predicted_num_purchases, t2.num_purchases) as rmse,
  mae(t1.predicted_num_purchases, t2.num_purchases) as mae
from
  predictions t1
join
  training t2 on (t1.id = t2.id)
;

Issue Links

• links to: GitHub Pull Request #210