Free online ARFF to XLSX converter

ARFF files are common in machine-learning workflows, but spreadsheet applications generally do not open them directly. Converting one to XLSX makes the records available for inspection, filtering, editing, and sharing in Excel or LibreOffice Calc.

What the ARFF format is

ARFF (Attribute-Relation File Format) is a text-based dataset format associated with the Weka machine-learning toolkit. An ARFF file contains a @relation declaration, @attribute definitions describing column names and types, and a @data section containing dense or sparse records.

Weka, KEEL, machine-learning research software, and custom Python or Java programs create ARFF files. You usually encounter them in classification, clustering, benchmarking, and educational datasets. ARFF supports numeric, nominal, string, date, and missing values represented by ?.

What the XLSX format is

XLSX is the XML-based spreadsheet format used by Microsoft Excel and supported by LibreOffice Calc, Google Sheets import, and other spreadsheet programs. It stores tabular values in worksheets and supports filtering, formulas, charts, formatting, and multiple sheets.

XLSX is preferable when people need to review or edit the dataset manually, send it to users without Weka, archive it with office documents, or perform spreadsheet-based analysis. The conversion normally places ARFF attributes in the first row and data records in the rows below.

How to actually convert ARFF to XLSX

Using Python

This method preserves the ARFF data table and handles nominal and sparse records through the liac-arff parser.

1. Install the required packages:

python -m pip install liac-arff pandas openpyxl

2. Save the following as arff_to_xlsx.py:

import arff
import pandas as pd

with open('input.arff', 'r', encoding='utf-8') as source:
    dataset = arff.load(source)

columns = [attribute[0] for attribute in dataset['attributes']]
frame = pd.DataFrame(dataset['data'], columns=columns)
frame.to_excel('output.xlsx', index=False, engine='openpyxl')

3. Put input.arff beside the script and run:

python arff_to_xlsx.py

The resulting output.xlsx contains one worksheet with the attribute names as column headers. To avoid Excel interpreting date-like or identifier values incorrectly, inspect the resulting columns and apply the required number formats after opening the workbook.

Using Weka and LibreOffice

Weka can first export the ARFF data to CSV, after which LibreOffice Calc can save it as XLSX:

java -cp weka.jar weka.core.converters.CSVSaver -i input.arff -o output.csv

Open output.csv in LibreOffice Calc, verify the delimiter and character encoding in the import dialog, then choose File → Save As, select Excel 2007–365 (*.xlsx), and save the workbook. This route is useful when Weka is already installed or when you want to inspect the exported CSV before creating the spreadsheet.

Using the online converter

For a conversion without installing software, upload the ARFF file to the free converter on 101convert.com and select XLSX as the output format. Download the generated workbook and check its columns and missing values before using it for analysis.

Quality and compatibility limits

XLSX does not have ARFF’s relation and attribute-declaration model. The relation name, nominal-value lists, date-format declarations, comments, and machine-learning metadata may not appear in the workbook unless they are added to a separate worksheet or preserved in the original ARFF file.

Missing values may remain as ?, become blank cells, or be converted to a library-specific missing value. Confirm the result before importing it into another program. Nominal attributes are exported as ordinary cell values; XLSX does not automatically retain their allowed-value lists.

Excel may automatically reinterpret long integers, leading-zero identifiers, dates, and scientific notation. Import those columns as text or apply an explicit format when their exact character representation matters. XLSX also has a limit of 1,048,576 rows and 16,384 columns per worksheet; larger ARFF datasets require splitting or another data format.