Creating a Report by Uploading a CSV to Data Sandbox

Self Service Report from CSV

15. Delete rows with empty or invalid values in relevant columns after applying the required transformations.

20. Save the data preparation and exit.

25. Plot a KPI to compare CTC by individuals using Name, Previous CTC, and Offered CTC.

31. Design a pie chart by selecting Candidate ID and Source, changing the style to doughnut, and adjusting the legend.

Prerequisites:

• CSV file (e.g., Hiring Data).

• Access to Data Center (Data Sandbox, Data Preparation, Data Store) and Report Module.

• Permission to create/save assets.

Phase 1 — Upload the CSV to Data Sandbox

1. Download the sample dataset (Hiring Data).

2. Open Apps Menu → Data Center.

3. Go to My Connectors → Data Sandbox.

4. Click Create, enter a Data Sandbox Name and optional Description.

   
5. Upload your CSV (from local machine) and wait for the upload to complete.

   • The file name appears at the bottom when successfully uploaded from the system.

   • Click the Upload option to complete the process.

   

6. A notification appears to inform that the selected file has been uploaded.

7. The recently created Sandbox entry will be listed at the top of the Data Sandbox list.

8. Click the Vertical Ellipsis icon to open the Options context menu for the newly uploaded Sandbox file.

9. Click Preview to verify the content.

   

Phase 2 — Clean the data with Data Preparation

1. In Data Sandbox, click the ⋯ (More) menu for the uploaded file → Create Data Preparation.

   • Data opens in a grid (rows/columns) with summary details below.

2. Profile the data and note issues:

   • Gender has 6 categories (expect 2) with 17 invalid rows.

   • Experience contains decimals (invalid format).

   • Previous CTC, Offered CTC, and Monthly Salary have empty rows (highlighted light blue).

3. Click AutoPrep and accept the default transformations.

   • Ensure special characters are removed from column headers.

4. Apply corrective steps:

   • Delete rows with empty/invalid values where appropriate.

   • Normalize Experience (fix decimal issues).

   • Impute blanks in Previous CTC, Offered CTC, and Monthly Salary with 0.

   • Validate changes in preview.

5. Reduce Gender categories:

   • After AutoPrep, Gender shows 4 categories.

   • Use Search & Replace to standardize to Male and Female (2 categories total).

6. Name the preparation (e.g., Hiring_dataprep) → Save and Exit.

   

Phase 3 — Create a Data Store

1. From the sandbox file’s ⋯ (More) menu, choose Create Data Store.

2. Enter a Data Store Name, select the prepared data (from Phase 2).

   
3. Validate and Save; wait for the creation to complete.

   • The Data Store now appears for reporting.

Note: Upload-based Data Stores won’t support some actions (e.g., View/Edit, Refresh Data). Use them primarily as reporting sources.

Phase 4 — Build a Self-Service Report

1. Open Apps Menu → Report Module.

2. Create a new Self Service Report using the Data Store created above.

3. Design visuals using fields under the Data Store:

   • KPI / Chart: Compare CTC by individual: use Name, Previous CTC, and Offered CTC.

   • Sort in descending order and limit to 6 entries.

   • Add Title → Save View.

   
4. Add tabs for additional views:

   • Bar/Column: Team vs Monthly Salary (descending by salary).

     • Configure chart properties, labels, title → Save View.

   • Doughnut (Pie style): Candidate ID by Source.

     • Adjust legend, titles, display names → Save View.

   • Treemap: team-wise candidate details using Team and Candidate ID.

     • Customize colors, titles, display names → Save View.

Phase 5 — Accelerate with NLP & Synonyms

• Use NLP search in the Report Module to create charts by typing queries (e.g., “monthly salary by skills”).

• Pin selected NLP charts to the report board; Save adjustments.

• In the Data Store, add Synonyms (e.g., map “salary” → Monthly Salary) to improve NLP recall.

• Re-run queries using synonyms, pin the best results, and save.

• A detailed self-service report gets created.

  

Outcome

• CSV data is uploaded, cleaned, standardized, and materialized as a Data Store.

• A multi-tab Self Service Report with KPIs and charts is published for analysis and sharing.

Tips & Good Practices

• Validate at each stage (prep and store) to catch issues early.

• Use AutoPrep to jump-start cleaning; then add targeted transforms (search/replace, impute, delete invalids).

• Standardize categories (e.g., Gender) to avoid fragmented segments.

• Keep field names clean (no special chars) for smoother reporting and NLP.

• For large CSVs, iterate on a sample in Data Preparation, then finalize on the full data.

Troubleshooting

• Unexpected categories (e.g., Gender): Use Search/Replace + value normalization in Data Preparation.

• Invalid numeric/date fields: Cast types and replace non-conforming values (or delete invalid rows).

• Report shows blank/duplicates: Recheck Data Store source (correct prep version) and chart field mappings.

• NLP misses terms: Add Synonyms to the Data Store and retry the query.