Goal → steps → outcome

How to use it in real work.

Every case tells you where to go, what to select and what you get. Start without SQL, then preserve repeatable work as a query.

Explore 8 workflows Open full manual

Executable guides

Eight common problems, explained.

Names in bold match tabs or actions visible in the application.

Data quality

Validate a Parquet file before reporting

Goal: find nulls, duplicates, mixed types and outliers, then reach the affected rows.

Use Open file and select the .parquet.
In Schema, review names, types and overall size.
In Quality, run checks and select each warning.
Open affected records in Data & SQL; use Profile to confirm distributions and common values.

Outcome: an investigable anomaly list and an optional HTML profile for the review file.

Manual: quality and profiling →

ETL and audit

Compare two dataset versions

Goal: check whether a new delivery changed rows, columns, types or nulls.

Open the old version, then use Add file as table for the new one.
Rename tables in the tree, for example old and new.
In Compare, select both tables and run the comparison.
For key-level checks, open Data & SQL and run a JOIN or ANTI JOIN.

Outcome: structural and content metrics plus SQL to locate added or missing records.

Manual: compare datasets →

No-code analysis

Build an operational summary and pivot

Goal: get an initial reading and answer “how much by category and period” without coding.

Open the file and use Dashboard to review distributions, categories and time series.
In Pivot, place a period or zone in rows and a category in columns.
Select a numeric measure and aggregation: sum, average, count, minimum or maximum.
Export to Excel or keep the generated query.

Outcome: a visual summary and tabular matrix ready for reporting or review.

Manual: dashboard and pivot →

Data science

Explore relationships before modeling

Goal: recognize related variables, patterns and possible non-linearities before moving to Python or R.

In Profile, confirm types, spread and extreme values.
In Correlation, calculate Pearson or Spearman and review the strongest pairs.
Open the pair detail as a scatter plot.
In X-Y Chart, adjust axes, grouping, sample and chart type.

Outcome: visual hypotheses and priority pairs for later statistical analysis.

Manual: correlation and charts →

Governance

Create a reusable data dictionary

Goal: document the meaning, unit and notes for every field inside the project.

Review names and types in Schema.
Open Dictionary and select the active table.
Complete description, unit, domain or notes per column.
Save the .pvpw workspace and export the dictionary to share it.

Outcome: metadata that travels with the project and reduces ambiguity between teams.

Manual: data dictionary →

GIS, fisheries and oceanography

Map coordinates and filter an area

Goal: view points or density, overlay a layer and send a geographic selection to results.

In Map, assign latitude and longitude and choose points or a density grid.
Select a weighting variable and time range if needed.
Add a local layer for zones, ports or boundaries.
Draw a box, lasso, circle or polygon; send the filter to Data & SQL and export GeoJSON or PNG.

Outcome: a reproducible spatial subset and a map with legend, scale and attribution.

Manual: map and selection →

Databases

Analyze an Oracle query locally

Goal: retrieve only the required result, combine it with files and send an output back when needed.

Choose Import from Oracle and enter server, service and user; the password is not stored.
Write a query limited to the columns and period needed.
Import the result as a workspace table and combine it with CSV, Excel or Parquet.
Export to a file or, with suitable permissions, use Export to Oracle.

Outcome: local analysis without Oracle Instant Client and with better-controlled credentials.

Manual: Oracle →

Interoperability

Open difficult formats and deliver the right one

Goal: recover RDS/RData, legacy-encoded CSV, complex Excel or nested JSON.

Open the file normally; if it contains multiple sheets or tables, select those needed.
Check structure in Schema and use the cell viewer for nested lists or JSON.
Filter or transform in Data & SQL.
Use Export: Parquet for types, Excel/CSV for exchange, SQLite for a portable database or GeoJSON for GIS.

Outcome: readable, typed data delivered independently of the originating program.

Manual: formats and exports →

Feature coverage

All 12 modules take part in these workflows.

Data & SQL, Schema, Statistics, Quality, Profile, Dashboard, Pivot, X-Y Chart, Correlation, Compare, Dictionary and Map share the same active table and context.

Explore features by module Download 4.60.3