Loading Data Lab…

Data Lab

Data Lab at /datalab is UtilBloom’s spreadsheet-oriented workspace for CSV and Excel files you would rather not upload to a cloud BI trial. Import a table, inspect column types and missing values, apply light cleaning steps, run descriptive stats, and chart results while the grid stays in browser memory. The workflow suits analysts validating a client export, students exploring a lab dataset, and founders sanity-checking metrics before they paste slides into a deck. Pair it with Charts Studio when you want publication-style figures without re-importing, or jump from Graph Lab when you need to plot a regression line you computed here. Nothing in the default path sends your raw file to UtilBloom servers.

What you can do

  • CSV and Excel import

    Load comma-separated files and Excel workbooks into an in-tab grid, preserving enough structure to inspect headers, sheet names, and column widths before you commit to transforms.

  • Column profiling

    See inferred types, null counts, distinct values, and simple distribution hints so you catch swapped date columns, trailing spaces, or numeric fields stored as text early.

  • Cleaning helpers

    Trim whitespace, normalize delimiters, filter rows, and patch obvious formatting issues without standing up a Python notebook for a ten-minute hygiene pass.

  • Descriptive statistics

    Summarize numeric columns with means, quantiles, and counts suitable for lab write-ups or stakeholder emails when a full warehouse query would be overkill.

  • In-lab charting

    Visualize filtered subsets directly from the grid so you can confirm a trend before exporting a cleaned CSV or opening Charts Studio for finer styling.

  • Export-minded workflow

    Copy or download sanitized tables when you are satisfied, keeping provenance in your browser history instead of a third-party dataset bucket.

Worked examples

Survey export with blank rows

responses.csv (12 columns, mixed text and 1–5 ratings)

Profile nulls in the rating columns, drop rows where consent=false, and compare before/after row counts in the summary panel before charting satisfaction by cohort.

Monthly revenue sheet

revenue_Q3.xlsx, sheet ‘Actuals’

Confirm the date column parsed as dates, cast currency strings to numbers, then plot month-over-month totals to spot a single outlier month worth auditing in your ERP.

Sensor log trim

device_log.csv with timestamp, temperature, humidity

Filter to a 48-hour window, compute basic stats on temperature, and export a smaller CSV for sharing with a vendor under NDA without uploading the full multi-gig archive.

How it compares

Cloud spreadsheet add-ons and notebook hosts often assume you will upload the dataset to their infrastructure. Data Lab stays closer to UtilBloom’s local-first tool library: the file is read in your tab, and you choose when a derivative leaves the machine. It is not a replacement for warehouse SQL, dbt, or team-wide Looker dashboards, those tools win on scheduled pipelines and governed metrics. Data Lab wins when you have one messy export, need profiling and charts in minutes, and already trust UtilBloom for PDF and math labs on the same domain. For chart polish without the full grid toolkit, open /charts. For scripting-heavy transforms, /codelab runs Python and SQL locally when you outgrow click-based cleaning.

Questions

Where does my CSV live while I work in Data Lab?

In browser memory on your device. Closing the tab discards unsaved state unless you export or rely on any local persistence the workspace exposes for drafts.

How large a file can Data Lab handle?

RAM and JavaScript engine limits dominate. Multi-million-row warehouses should be sampled in SQL or Python first; Data Lab shines on megabyte-scale exports typical of spreadsheets.

Does Data Lab replace Charts Studio?

They overlap on visualization, but Data Lab centers the grid, profiling, and cleaning. Charts Studio emphasizes figure types and styling when the table is already tidy.

Can I join two spreadsheets?

Light merges and filters are in scope; complex star-schema joins belong in Code Lab or an external database. Import one table at a time when exploring.

Will Excel macros run here?

No. Only cell values and sheet structure import. Recreate logic with in-lab transforms or export to Code Lab for scripted equivalents.

How do I document cleaning steps for a team?

Note the sequence you applied and export the result with a dated filename. For reproducible pipelines, mirror the same steps in version-controlled scripts.

Related tools

Guides