repository·master·Indexed 21 days ago
https://github.com/danielhrisca/asammdfA high-performance Python library for parsing, editing, and converting ASAM MDF (Measurement Data Format) files. It supports MDF versions 2, 3, and 4, providing a programmatic API and a graphical user interface for automotive data analysis. Key features include signal filtering, time-interval cutting, version conversion, and support for CAN/LIN bus logging via canmatrix.
asammdf is a high-performance parser and editor designed for ASAM (Association for Standardization of Automation and Measuring Systems) MDF (Measurement Data Format) files. It allows developers to programmatically read, edit, and manipulate measurement data.
.dat files.mdf files.mf4 filesThe asammdf library is designed for high-performance manipulation of MDF (Measurement Data Format) files.
Core Capabilities:
.dbc or .arxml databases), and filter subsets of channels.Signal class for unified operations across v3 and v4 files. This is preferred over pandas DataFrames when dealing with channels from different sources at different sampling rates.pandas, HDF5, Matlab (v7.3), CSV, and parquet.asammdf.blocks.v2_v3_blocks module contains classes that implement the various data blocks used in the MDF version 3 specification. These classes represent the structural components of an MDF file, such as channel definitions, data groups, headers, and identification blocks. While these are low-level components used for version conversion and file parsing, they define how the library handles the internal organization of MDFv3 files.The asammdf.mdf.MDF class delegates all attribute access to an underlying _mdf attribute, which is an instance of MDF2, MDF3, or MDF4 depending on the file version.
MDF3 instance (MDF2 and MDF3 share the same implementation).MDF4 instance.Users should refer to the MDF3 or MDF4 documentation for specific methods available to those versions.
When defining functions for virtual channels, you must follow these specific Python syntax rules:
def Function1(a, b, t=0)def Function1(a=1.5, b=0, t=0)t=0. This argument is reserved for the time stamp and cannot be used for other purposes.None or strings will cause errors.math as mathpandas as pdnumpy as npdef Function1(a=0, t=0):
if a > 5:
return 1
else:
return 0
# Function2 can call Function1
def Function2(b=1, c=7, t=0):
if b != 0:
return Function1(c)
else:
return c / 2When adding a virtual channel to a Plot window (shortcut Ins), you must choose a Computation mode. This determines how the Python function is executed:
t.if/else) or simple arithmetic.numpy ndarray for each argument, along with the full time base array t.numpy (e.g., np.clip, np.diff).asammdf.blocks.v4_blocks module contains classes that implement the various data structures (blocks) defined in the MDF version 4 specification. These classes represent the low-level components of an MDF file, such as headers, channel definitions, data groups, and attachment information. While most users interact with the high-level ASAMMFD object, these block classes are available for low-level manipulation or inspection of the MDF file structure.The GUI operates in three distinct modes that change how files are handled:
MDF methods is decoupled from the original file. Modifying a Signal object returned by a get call does not affect the raw data in the file. Subsequent calls to get for the same channel will always return the original data from the file.When initializing an asammdf MDF object, you can control how much data is loaded into RAM using the memory parameter. This allows you to balance between processing speed and memory consumption:
full: Everything is loaded into RAM. This provides the fastest access to channel data but consumes the most memory.low: Metadata is loaded into RAM, but raw channel data is not. This significantly reduces memory usage at the cost of slower data access (as data must be read from disk when requested).minimum: Uses the full loading strategy but aims for the lowest possible RAM footprint during the object creation process.The Single files page allows you to open multiple files individually for visualization or processing (e.g., exporting to CSV or HDF5).
Key features include:
Ctrl+F to open the advanced search dialog. It supports wildcard and regex patterns. The "Pattern based window" tab allows filtering channels by name pattern and signal value conditions.Shortcut: Ctrl+F (Advanced Search)Numeric windows are designed to handle large numbers of channels and are used to inspect instantaneous signal values or search for specific values.
Key Capabilities: