Common Import Errors

What goes wrong during file import, what you will see, and how to fix it.

Most import problems fall into one of four categories: the file itself is rejected before parsing begins, the file parses but produces unexpected column types, the structure of the data causes problems that only appear later during analysis, or a schema mismatch is detected when adding data to an existing project.

Once these issues are resolved, Licklider can move on to setup, quality checks, figures, and results using the dataset structure you intended to provide.

The file is rejected at upload

These errors appear immediately when you submit a file.

File too large Files over 25 MB are rejected. Split the file or reduce the number of rows before uploading.

Wrong file type Only .csv and .xlsx files are accepted. PDF, Word, PowerPoint, and image files are not data sources and cannot be imported.

Empty file A file with no content is rejected. Verify the file saved correctly before uploading.

The file uploads but columns look wrong

These problems do not produce an error message — the file appears to import successfully, but the column types or values in the data summary look unexpected.

Numeric column treated as categorical

This happens when a column that should be numeric contains non-numeric values. Common causes:

  • Cells with text like N/A, < 0.01, n.d., or ND mixed into a numeric column
  • A unit label embedded in the value, such as 42 mg/dL instead of 42
  • A note or flag appended to some cells, such as 5* or 12†

Fix: replace non-numeric entries with a blank cell, or standardize them before importing. Numeric columns must contain only numbers and empty cells to be recognized as numeric.

Temporal column treated as categorical or numeric

Licklider recognizes a column as temporal when most values parse as dates or time points. If a significant fraction of values are missing or in an unrecognized format, the column may fall back to categorical.

Fix: ensure date values are consistently formatted. ISO format (YYYY-MM-DD) is the most reliable. Remove or standardize any values that deviate from the main format.

All columns appear as categorical

This typically means the file contains multiple tables, a header row mixed into the data, or a large number of non-numeric entries. When the structural integrity of the table breaks down, type inference across all columns can fail at once.

See the structural problems section below.

Delimiter detected incorrectly

If your file uses a semicolon or tab as a delimiter instead of a comma, Licklider may read the entire row as a single column. The file will appear to import with one column containing the full text of each row.

Fix: open the file in a text editor or spreadsheet application and confirm the delimiter. Resave as a comma-separated CSV or change the delimiter to a comma before importing.

Two columns collapse to the same name

Licklider normalizes column names on import by removing invisible characters and trimming whitespace. If two columns produce the same name after normalization, both are retained but a collision warning is recorded. This may cause unexpected behavior in analyses that reference those columns.

Fix: rename columns to be distinct before importing.

Character encoding problems

Licklider automatically detects and converts UTF-8, UTF-8 with BOM, Shift-JIS (CP932), and EUC-JP. For other encodings, the file is read as UTF-8, which may produce garbled column names or values.

Column names or values appear as garbled characters

This typically means the file uses an encoding that was not detected correctly.

Fix: re-save the file as UTF-8 in your spreadsheet application before importing. In Excel, use File → Save As → CSV UTF-8.

Structural problems

These problems come from how the file is organized, not from individual values.

Multiple tables in one file

If your file contains more than one table — for example, a results block followed by a summary block, separated by blank rows — Licklider will read them as a single continuous table. Because blank rows are skipped during parsing, the header of the second table appears as a data row inside the first table. This causes type inference to fail across many columns.

Fix: remove all tables except the one you intend to analyze. Export each table as a separate file.

Title rows or notes above the header

If your file has a title, experiment name, or metadata note in the rows above the column headers, the first data row will be misread as the header. Column names will appear as the title text rather than your actual variable names.

Fix: delete all rows above the column header before importing. The first row must contain column names and nothing else.

Wide format when long format is needed

Wide-format files import without error. The file appears to load correctly, and column names are visible. However, when you request an analysis that requires a group column and a value column — such as a t-test or ANOVA — Licklider will report that the required columns are missing.

This is not a file error. It is a structural mismatch between the data layout and the analysis. See Table Shape: Wide vs Long for how to identify and fix this. In other words, the parser may have succeeded while the analysis structure is still not ready.

Schema drift

If you upload a file to a project that already has data, Licklider checks whether the new file matches the structure of the existing dataset. If the column count or column names differ, a schema drift warning or error is shown.

  • Warning: the mismatch is detected but the upload is allowed to proceed. The warning appears alongside the upload confirmation.
  • Error: the mismatch is severe enough that the upload is blocked. Correct the file structure to match the existing dataset before retrying.

XLSX-specific problems

Encrypted or password-protected files Licklider cannot read encrypted XLSX files. Remove the password protection before uploading.

File appears corrupt If the XLSX file fails to open, it may have been saved incorrectly or become corrupt during transfer. Try reopening it in Excel and saving again before uploading.

File is too large XLSX files are subject to the same 25 MB limit as CSV files. For large datasets, export as CSV instead.

If import behavior still looks unusual after addressing the above, re-save the sheet as a UTF-8 CSV and retry. This eliminates hidden formatting, merged cells, and encoding issues that can affect XLSX parsing.

If the problem is not listed here

If your file passes all the checks above but the data still does not look right after import, the most reliable next step is to open the file in a spreadsheet application, check the data summary that Licklider shows after upload, and compare the two. Column types, row counts, and value ranges are shown in the data summary and will usually reveal where the interpretation diverged.

If the file imports but the dataset still does not behave the way you expect in analysis setup, continue with Required and Optional Columns and ID, Batch, and Timepoint Columns.

What this page does not cover