Dataset

Attribute Information:
   1. sepal length in cm
   2. sepal width in cm
   3. petal length in cm
   4. petal width in cm
   5. class: 
      -- Iris Setosa
      -- Iris Versicolour
      -- Iris Virginica

Table preparation

create database iris;
use iris;

create external table raw (
  sepal_length int,
  sepal_width int,
  petal_length int,
  petak_width int,
  class string
)
ROW FORMAT DELIMITED
  FIELDS TERMINATED BY ','
  LINES TERMINATED BY '\n'
STORED AS TEXTFILE LOCATION '/dataset/iris/raw';

$ sed '/^$/d' iris.data | hadoop fs -put - /dataset/iris/raw/iris.data
create table label_mapping 
as
select
  class,
  rank - 1 as label
from (
select
  distinct class,
  dense_rank() over (order by class) as rank
from 
  raw
) t
;
create table training
as
select
  rowid() as rowid,
  array(t1.sepal_length, t1.sepal_width, t1.petal_length, t1.petak_width) as features,
  t2.label
from
  raw t1
  JOIN label_mapping t2 ON (t1.class = t2.class)
;

Training

train_randomforest_classifier takes a dense features in double[] and a label starting from 0.

CREATE TABLE model 
  STORED AS SEQUENCEFILE 
AS
select 
  train_randomforest_classifier(features, label) 
  -- hivemall v0.4.1-alpha.2 and before
  -- train_randomforest_classifier(features, label) as (pred_model, var_importance, oob_errors, oob_tests)
  -- hivemall v0.4.1 and later
  -- train_randomforest_classifier(features, label) as (model_id, model_type, pred_model, var_importance, oob_errors, oob_tests)
from
  training;

Caution

The default TEXTFILE should not be used for model table when using Javascript output through -output javascript option.

hive> desc extended model;
col_name data_type
model_id string
model_weight double
model string
var_importance array
oob_errors int
oob_tests int

Training options

-help option shows usage of the function.

select train_randomforest_classifier(features, label, "-help") from training;

> FAILED: UDFArgumentException 
usage: train_randomforest_classifier(array<double|string> features, int
       label [, const array<double> classWeights, const string options]) -
       Returns a relation consists of <int model_id, int model_type,
       string pred_model, array<double> var_importance, int oob_errors,
       int oob_tests, double weight> [-attrs <arg>] [-depth <arg>] [-help]
       [-leafs <arg>] [-min_samples_leaf <arg>] [-rule <arg>] [-seed
       <arg>] [-splits <arg>] [-stratified] [-subsample <arg>] [-trees
       <arg>] [-vars <arg>]
 -attrs,--attribute_types <arg>      Comma separated attribute types (Q
                                     for quantitative variable and C for
                                     categorical variable. e.g.,
                                     [Q,C,Q,C])
 -depth,--max_depth <arg>            The maximum number of the tree depth
                                     [default: Integer.MAX_VALUE]
 -help                               Show function help
 -leafs,--max_leaf_nodes <arg>       The maximum number of leaf nodes
                                     [default: Integer.MAX_VALUE]
 -min_samples_leaf <arg>             The minimum number of samples in a
                                     leaf node [default: 1]
 -rule,--split_rule <arg>            Split algorithm [default: GINI,
                                     ENTROPY]
 -seed <arg>                         seed value in long [default: -1
                                     (random)]
 -splits,--min_split <arg>           A node that has greater than or
                                     equals to `min_split` examples will
                                     split [default: 2]
 -stratified,--stratified_sampling   Enable Stratified sampling for
                                     unbalanced data
 -subsample <arg>                    Sampling rate in range (0.0,1.0]
 -trees,--num_trees <arg>            The number of trees for each task
                                     [default: 50]
 -vars,--num_variables <arg>         The number of random selected
                                     features [default:
                                     ceil(sqrt(x[0].length))].
                                     int(num_variables * x[0].length) is
                                     considered if num_variable is (0,1

Caution

-num_trees controls the number of trees for each task, not the total number of trees.

Parallelize Training

To parallelize RandomForest training, you can use UNION ALL as follows:

CREATE TABLE model 
  STORED AS ORC tblproperties("orc.compress"="SNAPPY")
  -- STORED AS SEQUENCEFILE 
AS
select 
  train_randomforest_classifier(features, label, '-trees 25') 
from
  training
UNION ALL
select 
  train_randomforest_classifier(features, label, '-trees 25')
from
  training
;

Learning stats

Variable importance and Out Of Bag (OOB) error rate of RandomForest can be shown as follows:

select
  array_sum(var_importance) as var_importance,
  sum(oob_errors) / sum(oob_tests) as oob_err_rate
from
  model;

[6.837674865013268,4.1317115752776665,24.331571871930226,25.677497925673062] 0.056666666666666664

Prediction

set hivevar:classification=true;
set hive.auto.convert.join=true;
set hive.mapjoin.optimized.hashtable=false;

create table predicted
as
SELECT
  rowid,
  -- rf_ensemble(predicted) as predicted
  -- hivemall v0.5-rc.1 or later
  rf_ensemble(predicted.value, predicted.posteriori, model_weight) as predicted
  -- rf_ensemble(predicted.value, predicted.posteriori) as predicted -- avoid OOB accuracy (i.e., model_weight)
FROM (
  SELECT
    rowid, 
    -- hivemall v0.4.1 and later
    -- tree_predict(p.model_id, p.model_type, p.pred_model, t.features, ${classification}) as predicted
    -- hivemall v0.5-rc.1 or later
    p.model_weight,
    tree_predict(p.model_id, p.model, t.features, ${classification}) as predicted
    -- tree_predict_v1(p.model_id, p.model_type, p.pred_model, t.features, ${classification}) as predicted -- to use the old model in v0.5-rc.1 or later
  FROM
    model p
    LEFT OUTER JOIN -- CROSS JOIN
    training t
) t1
group by
  rowid
;

Caution

tree_predict_v1 is for the backward compatibility for using prediction models built before v0.5-rc.1 on v0.5-rc.1 or later.

Parallelize Prediction

The following query runs predictions in N-parallel. It would reduce elapsed time for prediction almost by N.

SET hivevar:classification=true;
set hive.auto.convert.join=true;
SET hive.mapjoin.optimized.hashtable=false;
SET mapred.reduce.tasks=8;

create table predicted
as
SELECT
  rowid,
  -- rf_ensemble(predicted) as predicted
  -- hivemall v0.5-rc.1 or later
  rf_ensemble(predicted.value, predicted.posteriori, model_weight) as predicted
  -- rf_ensemble(predicted.value, predicted.posteriori) as predicted -- avoid OOB accuracy (i.e., model_weight)
FROM (
  SELECT
    t.rowid, 
    -- hivemall v0.4.1 and later
    -- tree_predict(p.model_id, p.model_type, p.pred_model, t.features, ${classification}) as predicted
    -- hivemall v0.5-rc.1 or later
    p.model_weight,
    tree_predict(p.model_id, p.model, t.features, ${classification}) as predicted
    -- tree_predict_v1(p.model_id, p.model_type, p.pred_model, t.features, ${classification}) as predicted as predicted -- to use the old model in v0.5-rc.1 or later
  FROM (
    SELECT 
      -- hivemall v0.4.1 and later
      -- model_id, model_type, pred_model
      -- hivemall v0.5-rc.1 or later
      model_id, model_weight, model
    FROM model
    DISTRIBUTE BY rand(1)
  ) p 
  LEFT OUTER JOIN training t
) t1
group by
  rowid;

Evaluation

select count(1) from training;

150

set hivevar:total_cnt=150;

WITH t1 as (
SELECT
  t.rowid,
  t.label as actual,
  p.predicted.label as predicted
FROM
  predicted p
  LEFT OUTER JOIN training t ON (t.rowid = p.rowid)
)
SELECT
  count(1) / ${total_cnt}
FROM
  t1
WHERE
  actual = predicted
;

0.98

Graphvis export

Note

tree_export feature is supported from Hivemall v0.5-rc.1 or later. Better to limit tree depth on training by -depth option to plot a Decision Tree.

Hivemall provide tree_export to export a decision tree into Graphviz or human-readable Javascript format. You can find the usage by issuing the following query:

> select tree_export("","-help");

usage: tree_export(string model, const string options, optional
       array<string> featureNames=null, optional array<string>
       classNames=null) - exports a Decision Tree model as javascript/dot]
       [-help] [-output_name <arg>] [-r] [-t <arg>]
 -help                             Show function help
 -output_name,--outputName <arg>   output name [default: predicted]
 -r,--regression                   Is regression tree or not
 -t,--type <arg>                   Type of output [default: js,
                                   javascript/js, graphvis/dot
CREATE TABLE model_exported 
  STORED AS ORC tblproperties("orc.compress"="SNAPPY")
AS
select
  model_id,
  tree_export(model, "-type javascript", array('sepal_length','sepal_width','petal_length','petak_width'), array('Setosa','Versicolour','Virginica')) as js,
  tree_export(model, "-type graphvis", array('sepal_length','sepal_width','petal_length','petak_width'), array('Setosa','Versicolour','Virginica')) as dot
from
  model
-- limit 1
;
digraph Tree {
 node [shape=box, style="filled, rounded", color="black", fontname=helvetica];
 edge [fontname=helvetica];
 0 [label=<petal_length &le; 2.599999964237213>, fillcolor="#00000000"];
 1 [label=<predicted = Setosa>, fillcolor="0.0000,1.000,1.000", shape=ellipse];
 0 -> 1 [labeldistance=2.5, labelangle=45, headlabel="True"];
 2 [label=<petal_length &le; 4.950000047683716>, fillcolor="#00000000"];
 0 -> 2 [labeldistance=2.5, labelangle=-45, headlabel="False"];
 3 [label=<petak_width &le; 1.6500000357627869>, fillcolor="#00000000"];
 2 -> 3;
 4 [label=<predicted = Versicolour>, fillcolor="0.3333,1.000,1.000", shape=ellipse];
 3 -> 4;
 5 [label=<sepal_width &le; 3.100000023841858>, fillcolor="#00000000"];
 3 -> 5;
 6 [label=<predicted = Virginica>, fillcolor="0.6667,1.000,1.000", shape=ellipse];
 5 -> 6;
 7 [label=<predicted = Versicolour>, fillcolor="0.3333,1.000,1.000", shape=ellipse];
 5 -> 7;
 8 [label=<petak_width &le; 1.75>, fillcolor="#00000000"];
 2 -> 8;
 9 [label=<petal_length &le; 5.299999952316284>, fillcolor="#00000000"];
 8 -> 9;
 10 [label=<predicted = Versicolour>, fillcolor="0.3333,1.000,1.000", shape=ellipse];
 9 -> 10;
 11 [label=<predicted = Virginica>, fillcolor="0.6667,1.000,1.000", shape=ellipse];
 9 -> 11;
 12 [label=<predicted = Virginica>, fillcolor="0.6667,1.000,1.000", shape=ellipse];
 8 -> 12;
}

Iris Graphvis output

You can draw a graph by dot -Tpng iris.dot -o iris.png or using Viz.js.

results matching ""

    No results matching ""