Technical Report: CSVM format for scientific tabular data

TL;DR

The paper introduces CSVM, a format based on CSV that includes metadata for better storage, exchange, and long-term use of scientific tabular data, compatible with common tools.

Contribution

It presents the first release of CSVM, a metadata-enriched extension of CSV for scientific data management and exchange.

Findings

CSVM files are readable by standard spreadsheet tools.

CSVM enhances data exchange with embedded metadata.

CSVM supports annotation within data blocks.

Abstract

The CSVM (CSV with metadata data) is issued from CSV format and used for storing experimental data, models, specifications. CSVM allows the storage of tabular data with a limited but extensible amount of metadata. This increases the exchange and long term use of RAW data because all information needed to use subsequently the data are included in the CSVM file. Basic CSVM files are readable by current tools (i.e. spreadsheets) for handling tables. Using full possibilities of concept, it is possible to deviate from a strict table and annotate also inside the data block. CSVM file are pure ASCII files and could provide a template for implementing best practices in handling raw data at a laboratory level, in exchange between data sources, in long term resources, or in collaborative processes particularly when different scientific fields are implied. In this document we describe the first (CSVM-1) release of CSVM format.