Supported File Formats

What Licklider accepts, what it expects inside the file, and how to avoid the most common format-level problems.

Licklider is designed for tabular experimental data. The accepted input is a single structured table — not a report, not a workbook, not an image.


Currently supported

If your file can be read cleanly at this stage, Licklider can move on to structure confirmation, analysis setup, figures, quality checks, and exportable results.

CSV (.csv)

PropertyRequirement
Extension.csv
EncodingUTF-8, UTF-8 with BOM, Shift-JIS (CP932), EUC-JP — auto-detected
DelimiterComma (,), tab (\t), or semicolon (;) — auto-detected; tab-separated .tsv extension not yet supported
Header rowRequired — first row must contain column names
StructureSingle table per file
Row conventionOne row = one observation or one record
Size limit25 MB (fast path) or up to 100 MB (Large Dataset Mode)

After upload, the behavior depends on file size. Files up to 25 MB are processed immediately and the dataset opens directly. Files between 25 MB and 100 MB enter Large Dataset Mode: the file is processed asynchronously, a sampled preview is shown while profiling runs, and the dataset surface opens once processing is complete. Licklider will ask clarifying questions if the structure is ambiguous — for example, if group columns or observation units are not immediately clear.

Excel (.xlsx)

.xlsx files are accepted as a compatibility path. Licklider reads the first sheet only. Additional sheets, charts, images, and macros are ignored. The size limit for Excel files is 25 MB. Large Dataset Mode, which supports files up to 100 MB, applies to CSV files only.

For best results, clean up the file before uploading: remove merged cells, decorative rows, and anything that is not part of the data table itself.


The format least likely to cause problems:

  • UTF-8 CSV, no BOM
  • First row contains column names, nothing else
  • No title rows, blank rows, or summary rows mixed into the data
  • One table per file — not multiple blocks separated by blank lines
  • No merged cells, colored headers, or formatting of any kind
  • Numbers stored as numbers, not as text with units embedded (e.g., 42 not "42 mg/dL")

If your file matches this description, upload will be smooth and Licklider's structure inference will be reliable. This is why CSV is the recommended path for research workflows: it reduces avoidable parsing ambiguity before downstream analysis begins.


Not supported

FormatStatus
Excel with multiple sheetsFirst sheet only; others are silently ignored
.xls (legacy Excel)Not supported — save as .xlsx or .csv first
PDF, Word, PowerPointNot accepted as data sources
Image files (TIFF, PNG, etc.)Not accepted
Real-time or streaming dataNot applicable
Tab-separated values (.tsv)Not yet supported — rename to .csv and confirm delimiter
CSV files larger than 100 MBNot supported
Excel (.xlsx) files larger than 25 MBNot supported
Report-style CSVsParsed as data — structure problems will surface during import

What Licklider expects inside the file

Licklider interprets your file as a rectangular table:

  • Each column represents one variable
  • Each row represents one observation, sample, or record
  • Values in a column are consistent in type — not a mix of numbers, units, and notes

You do not need to label columns in any special way at this stage. Licklider will ask about column roles — which column identifies groups, which is the outcome, which is an ID — during the setup step.

For details on column naming and required fields, see Required and Optional Columns.


Common format-level problems

These issues are caused by file structure before any analysis begins.

Large CSV files (over 25 MB) may enter a different processing flow with a sampled preview and asynchronous profiling before the full dataset becomes available.

ProblemWhat happens
Character encoding not detected correctlyColumn names or values appear garbled
Delimiter is semicolon or tab, not commaFile parses as a single column
Title or description rows before the headerColumn names are misread or skipped
Blank rows or blank columns in the middle of the tableTable is split or truncated
Multiple tables in one fileOnly the first contiguous block is read reliably

For a full list of import errors and how to fix them, see Common Import Errors.


If your file is not ready yet

  • No data at hand? Use the sample dataset on the start screen — it covers a typical group comparison workflow
  • Working in Excel? Export to .csv via File → Save As before uploading
  • Unsure how to structure your rows and columns? See Table Shape: Wide vs Long
  • Questions about which columns are required? See Required and Optional Columns
  • Want the bigger-picture overview of what happens after upload? See What This Product Does or Quickstart