mmlspark-101: TrainClassifier

mmlspark 101

預測一個人的收入是否超過$50k
數據下載地址https://www.kaggle.com/uciml/adult-census-income/data
注意!!!
mmlspark安裝,版本0.17,部分api已經發生變化,官方git上notebook版本較低
shell

pyspark --master=spark://Lord:7077 --packages Azure:mmlspark:0.17

會自動下載

from pyspark import SparkConf
import pyspark.sql.functions as F
from pyspark.sql import SparkSession
from pyspark.sql.types import StructType, StructField
from pyspark.mllib.evaluation import RankingMetrics, RegressionMetrics
from pyspark.sql.types import StringType, FloatType, IntegerType, LongType

read and clean data

spark = SparkSession.builder.appName("MyApp").config("spark.jars.packages", "Azure:mmlspark:0.17").getOrCreate()

data = spark.read.csv('hdfs:///user/hadoop/adult.csv',inferSchema=True, header=True)

data.limit(10).toPandas()

	age	workclass	fnlwgt	education	education.num	marital.status	occupation	relationship	race	sex	capital.gain	capital.loss	hours.per.week	native.country	income
0	90	?	77053	HS-grad	9	Widowed	?	Not-in-family	White	Female	0	4356	40	United-States	<=50K
1	82	Private	132870	HS-grad	9	Widowed	Exec-managerial	Not-in-family	White	Female	0	4356	18	United-States	<=50K
2	66	?	186061	Some-college	10	Widowed	?	Unmarried	Black	Female	0	4356	40	United-States	<=50K
3	54	Private	140359	7th-8th	4	Divorced	Machine-op-inspct	Unmarried	White	Female	0	3900	40	United-States	<=50K
4	41	Private	264663	Some-college	10	Separated	Prof-specialty	Own-child	White	Female	0	3900	40	United-States	<=50K
5	34	Private	216864	HS-grad	9	Divorced	Other-service	Unmarried	White	Female	0	3770	45	United-States	<=50K
6	38	Private	150601	10th	6	Separated	Adm-clerical	Unmarried	White	Male	0	3770	40	United-States	<=50K
7	74	State-gov	88638	Doctorate	16	Never-married	Prof-specialty	Other-relative	White	Female	0	3683	20	United-States	>50K
8	68	Federal-gov	422013	HS-grad	9	Divorced	Prof-specialty	Not-in-family	White	Female	0	3683	40	United-States	<=50K
9	41	Private	70037	Some-college	10	Never-married	Craft-repair	Unmarried	White	Male	0	3004	60	?	>50K

withColumnRenamed

data = data.withColumnRenamed('education.num','education_num')\
           .withColumnRenamed('marital.status','marital_status')\
           .withColumnRenamed('capital.gain','capital_gain')\
           .withColumnRenamed('capital.loss','capital_loss')\
           .withColumnRenamed('hours.per.week','hours_per_week')\
           .withColumnRenamed('native.country','native_country')

data.printSchema()

root
 |-- age: integer (nullable = true)
 |-- workclass: string (nullable = true)
 |-- fnlwgt: integer (nullable = true)
 |-- education: string (nullable = true)
 |-- education_num: integer (nullable = true)
 |-- marital_status: string (nullable = true)
 |-- occupation: string (nullable = true)
 |-- relationship: string (nullable = true)
 |-- race: string (nullable = true)
 |-- sex: string (nullable = true)
 |-- capital_gain: integer (nullable = true)
 |-- capital_loss: integer (nullable = true)
 |-- hours_per_week: integer (nullable = true)
 |-- native_country: string (nullable = true)
 |-- income: string (nullable = true)

EDA

import seaborn as sns
import matplotlib.pyplot as plt
%matplotlib inline

# add label column
add_label = F.udf(lambda income : 0 if income == '<=50K' else 1 , IntegerType())

data = data.withColumn('label', add_label(data['income']))

# 收入大於$50k
data[data.label == 1].describe().toPandas()[['summary','age','education_num','hours_per_week']]

	summary	age	education_num	hours_per_week
0	count	7841	7841	7841
1	mean	44.24984058155847	11.611656676444332	45.473026399693914
2	stddev	10.519027719851813	2.385128632665079	11.01297093020927
3	min	19	2	1
4	max	90	16	99

# 收入小於$50k
data[data.label == 0].describe().toPandas()[['summary','age','education_num','hours_per_week']]

	summary	age	education_num	hours_per_week
0	count	24720	24720	24720
1	mean	36.78373786407767	9.595064724919094	38.840210355987054
2	stddev	14.020088490824895	2.4361467923083993	12.31899464185489
3	min	17	1	1
4	max	90	16	99

age,education_num, hours_per_week分佈情況

ages_0 = data[data.label == 0].select('age').collect()

ages_1 = data[data.label == 1].select('age').collect()

plt.figure(figsize=(10, 5))
sns.distplot(ages_0, label='<=$50K')
sns.distplot(ages_1, label='>$50K')
plt.xlabel('age',fontsize=15)
plt.legend()

顯而易見,收入大於$50K$的人羣年齡整體大於小於$50K$

edus_0 = data[data.label == 0].select('education_num').collect()

edus_1 = data[data.label == 1].select('education_num').collect()

plt.figure(figsize=(10, 5))
sns.distplot(edus_0, label='<=$50K')
sns.distplot(edus_1, label='>$50K')
plt.xlabel('education_num',fontsize=15)
plt.legend()

hours_per_week_0 = data[data.label == 0].select('hours_per_week').collect()

hours_per_week_1 = data[data.label == 1].select('hours_per_week').collect()

plt.figure(figsize=(10, 5))
sns.distplot(hours_per_week_0, label='<=$50K')
sns.distplot(hours_per_week_1, label='>$50K')
plt.xlabel('hours_per_week',fontsize=15)
plt.legend()

Spilt Data and Training Data

data = data.select(["age","education", "education_num","marital_status", "hours_per_week", "income"])

train, test = data.randomSplit([0.75, 0.25], seed=20200420)

from mmlspark import TrainClassifier

from pyspark.ml.classification import LogisticRegression, DecisionTreeClassifier, NaiveBayes

# lr : LogisticRegression
# dt : decisionTree
# nb : NaiveBayes
model_lr = TrainClassifier(model=LogisticRegression(), labelCol="income", numFeatures=256).fit(train)
model_dt = TrainClassifier(model=DecisionTreeClassifier(), labelCol="income", numFeatures=256).fit(train)
model_nb = TrainClassifier(model=NaiveBayes(), labelCol="income", numFeatures=256).fit(train)

model_lr.write().overwrite().save("../models/LrModel.mml")

evaluate

from mmlspark import ComputeModelStatistics, TrainedClassifierModel
predictionModel = TrainedClassifierModel.load("../models/LrModel.mml")
prediction = predictionModel.transform(test)
metrics = ComputeModelStatistics().transform(prediction)
metrics.toPandas()

	evaluation_type	confusion_matrix	accuracy	precision	recall	AUC
0	Classification	DenseMatrix([[5766., 473.],\n [ 9...	0.824068	0.674242	0.503083	0.869809

dt_prediction = model_dt.transform(test)
nb_prediction = model_nb.transform(test)

dt_metrics = ComputeModelStatistics().transform(dt_prediction)
nb_metrics = ComputeModelStatistics().transform(nb_prediction)

dt_metrics.toPandas()

	evaluation_type	confusion_matrix	accuracy	precision	recall	AUC
0	Classification	DenseMatrix([[5943., 296.],\n [11...	0.825901	0.734052	0.419836	0.672863

nb_metrics.toPandas()

	evaluation_type	confusion_matrix	accuracy	precision	recall	AUC
0	Classification	DenseMatrix([[5824., 415.],\n [10...	0.818448	0.678295	0.44964	0.247985