Parsing Conventions and Capabilities
MagnetoPy can read in a variety of raw magnetic data and convert it to useful Python objects. The following sections describe what sort of file conventions are compatible with MagnetoPy and how the automatic parsing works.
Sample Info and Moment Scaling
When creating a Magnetometry
object, MagnetoPy will assume that all files within the given folder are from the same sample. Thus, the Magnetometry.sample_info
attribute will be created by looking at the sample information in the header of the first file that it finds.
The sample information is also used to automatically scale the magnetic moment data in all dc measurements (i.e., all MvsH
, ZFC
, and FC
objects within the dataset).
.dat File Conventions
MagnetoPy determines what experiment(s) is/are contained in a given file by checking the following items in this order:
- comments embedded in the data section of the file (e.g., the section of the file after the row containing "[DATA]" in a Quantum Design MPMP3 .dat file)
- the file name
- the actual data
Commented Files
In Quantum Design's MPMS3 MultiVu software, comments are added using the "Datafile Comment" section of the Sequence Command panel.
The comment should be placed immediately after setting the file location, as shown in the following example of a sequence containing ZFC and FC experiments.
Comments should be comma separated and can be in any order. The following experiments should be commented as follows:
MvsH
Comments must include:
- "MvsH" (case insensitive)
- the nominal temperature written as "XX", "XX K", or "XX C" (values without a unit are assumed to be K)
ZFCFC, ZFC, FC
Comments must include:
- "ZFC" or "FC" (case insensitive)
- the nominal magnetic field strength written as "XX", "XX Oe", or "XX T" (values without a unit are assumed to be Oe)
File Names
Recognized file names are (case insensitive):
- "zfcfc"
- further parsing will assume the file contain a single ZFC/FC experiment
- further parsing will assume the experiments are in the order ZFC then FC
- "zfc"
- further parsing will assume the file contain a single ZFC experiment
- "fc"
- further parsing will assume the file contain a single FC experiment
- "mvsh"
- further parsing can handle one or multiple MvsH experiments within the same file
Uncommented Files
Differentiating between files containing ZFCFC and MvsH experiments is done by checking the number of unique field values in the data, since it is assumed that ZFCFC files will only have a single field value.
ZFCFC, ZFC, FC
The automatic parsing of uncommented files containing ZFC/FC experiments currently only supports files with a single ZFC/FC pair. The data must be in the order ZFC then FC. The FC experiment can be collected with increasing or decreasing temperature.
MvsH
Files containing M vs. H data can contain one or multiple experiments at different nominal temperatures. Due to the temperature fluctuations, experiments are grouped by nominal temperatures using scikit-learn's DBSCAN clustering algorithm. Default values for the eps
and min_samples
parameters are provided and can be overwritten during MvsH
object creation or Magnetometry.extract_mvsh()