MLA-C01 Data Preparation for Machine Learning — All Questions With Answers

Refer to the exhibit. A data scientist runs the following AWS Glue ETL job script (Spark) to prepare data for ML:

```python
import sys
from awsglue.transforms import *
from awsglue.utils import getResolvedOptions
from pyspark.context import SparkContext
from awsglue.context import GlueContext
from awsglue.job import Job

args = getResolvedOptions(sys.argv, ['JOB_NAME'])
sc = SparkContext()
glueContext = GlueContext(sc)
spark = glueContext.spark_session
job = Job(glueContext)
job.init(args['JOB_NAME'], args)

datasource = glueContext.create_dynamic_frame.from_options(
    connection_type = "s3",
    connection_options = {"paths": ["s3://bucket/input/"]},
    format = "csv",
    format_options = {"withHeader": True}
)

applymapping = ApplyMapping.apply(frame = datasource, mappings = [("id", "int", "id", "int"), ("value", "string", "value", "double")])
...
```

The job fails with an error: "Job run failed: org.apache.spark.SparkException: Job aborted due to stage failure: Task failed while writing rows." What is the most likely cause of this error?

Refer to the exhibit. A data scientist is trying to run a SageMaker Processing job that reads data from an S3 bucket. The IAM role attached to the processing job has the following policy:

```json
{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "s3:GetObject",
            "Resource": "arn:aws:s3:::my-bucket/*"
        }
    ]
}
```

The job fails with an error: "Access Denied" when trying to list objects. What is the root cause?

Refer to the exhibit. A data scientist uses the following SageMaker Feature Store feature definition (using the Boto3 SDK) to create a feature group:

```python
import boto3
sagemaker = boto3.client('sagemaker', region_name='us-east-1')
response = sagemaker.create_feature_group(
    FeatureGroupName='my-feature-group',
    RecordIdentifierFeatureName='customer_id',
    EventTimeFeatureName='timestamp',
    FeatureDefinitions=[
        {'FeatureName': 'customer_id', 'FeatureType': 'String'},
        {'FeatureName': 'age', 'FeatureType': 'String'},
        {'FeatureName': 'income', 'FeatureType': 'Fractional'}
    ],
    OnlineStoreConfig={'EnableOnlineStore': True},
    RoleArn='arn:aws:iam::123456789012:role/SageMakerRole'
)
```

The data scientist later tries to ingest data with an 'age' column containing integer values. The ingestion fails. What is the most likely reason?

{
  "ProcessingResources": {
    "ClusterConfig": {
      "InstanceCount": 1,
      "InstanceType": "ml.m5.large",
      "VolumeSizeInGB": 30
    }
  },
  "AppSpecification": {
    "ImageUri": "123456789012.dkr.ecr.us-west-2.amazonaws.com/my-custom-image:latest",
    "ContainerEntrypoint": ["python", "process.py"]
  },
  "RoleArn": "arn:aws:iam::123456789012:role/SageMakerProcessingRole",
  "ProcessingInputs": [
    {
      "InputName": "input-1",
      "S3Input": {
        "S3Uri": "s3://my-bucket/input/data.csv",
        "LocalPath": "/opt/ml/processing/input",
        "S3DataType": "S3Prefix",
        "S3InputMode": "File",
        "S3DataDistributionType": "FullyReplicated",
        "S3CompressionType": "None"
      }
    }
  ],
  "ProcessingOutputConfig": {
    "Outputs": [
      {
        "OutputName": "output-1",
        "S3Output": {
          "S3Uri": "s3://my-bucket/output/",
          "LocalPath": "/opt/ml/processing/output",
          "S3UploadMode": "EndOfJob"
        }
      }
    ]
  }
}

An AWS Glue job fails with the following error from the CloudWatch logs:
"Conversion error: Unable to convert column 'price' from String to Double for some rows."

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:GetObject",
        "s3:PutObject"
      ],
      "Resource": "arn:aws:s3:::ml-data-bucket/*"
    }
  ]
}

import sys
from awsglue.transforms import *
from awsglue.utils import getResolvedOptions
from pyspark.context import SparkContext
from awsglue.context import GlueContext
from awsglue.job import Job

args = getResolvedOptions(sys.argv, ['JOB_NAME'])
sc = SparkContext()
glueContext = GlueContext(sc)
spark = glueContext.spark_session
job = Job(glueContext)
job.init(args['JOB_NAME'], args)

raw = glueContext.create_dynamic_frame.from_options(
    connection_type="s3",
    connection_options={"paths": ["s3://bucket/input/year=2023/month=01/"]},
    format="json")

transformed = raw.select_fields(["col1", "col2"]).rename_field("col1", "new_col")

glueContext.write_dynamic_frame.from_options(
    frame=transformed,
    connection_type="s3",
    connection_options={"path": "s3://bucket/output/transformed"},
    format="parquet")

job.commit()

ProcessingJobError: Execution failed
Error: Traceback (most recent call last):
  File "/opt/ml/processing/input/code/preprocess.py", line 45, in <module>
    df['age'] = df['age'].apply(float)
ValueError: could not convert string to float: 'twenty-five'