Prepare and Clean Datasets SNOWPRO-ADVANCED-DATA-ANALYST Practice Questions (Page 6)

Questions 26–30

1. 26foundation · easy

   Which SQL function in Snowflake converts a string to a numeric value?

   • ATO_CHAR
   • BCAST_AS_INT
   • CTO_NUMBER
   • DTO_DATE
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A retail analytics team loads a daily sales transactions table into Snowflake. Before building a revenue report, they run a profiling query and notice that the 'quantity' column contains the value -5 in 12 rows, and the 'unit_price' column contains 0 in 30 rows. The business rule states that quantity must be positive and unit_price must be greater than zero. What should the analyst do first to address these anomalies?

   • AFilter out the rows with negative quantity and zero unit_price from the report query.
   • BConvert the quantity and unit_price columns to VARCHAR to prevent further invalid numeric entries.
   • CReplace all negative quantities with NULL and all zero unit prices with the column average.
   • DFlag the rows in a separate quality-check table and notify the source system owner before deciding on correction or exclusion.
   Reveal Solution Discussion Select an answer first
4. 28foundation · easy

   What is feature engineering?

   • ADeleting features that are not needed
   • BVisualizing features in a chart
   • CStoring features in a separate database
   • DCreating new features or transforming existing ones to improve model performance
   Reveal Solution Discussion Select an answer first
5. 29expert · hard

   A data scientist is preparing a dataset of household incomes for a predictive model. The distribution is heavily right-skewed, and the IQR method flags several high-income households as outliers. The business context is that these high-income households are legitimate and important for the model's target audience. The model will use a linear regression algorithm. What is the most appropriate treatment?

   • ACap the income values at the 95th percentile to reduce their influence.
   • BRemove the flagged outliers to improve the model's fit.
   • CConvert income into a binary high/low category based on the median.
   • DApply a log transformation to the income variable.
   Reveal Solution Discussion Select an answer first
6. 30foundation · easy

   What is the purpose of creating a sample of data?

   • ATo remove all missing values from the dataset
   • BTo convert the data to a different format
   • CTo increase the size of the dataset
   • DTo create a smaller, representative subset for testing or analysis
   Reveal Solution Discussion Select an answer first