Skip to content

Releases: mHealthGroup/MIMSunit

MIMSunit 0.11.2

22 Jun 02:32
Compare
Choose a tag to compare

Remove “matlab” dependency.

BREAKING CHANGES

Remove “remove_average” and “bessel” filtering functions.

MIMSunit 0.11.1

24 Nov 03:33
Compare
Choose a tag to compare
v0.11.1

Fix CRAN check error

v0.11.0

04 Sep 03:04
Compare
Choose a tag to compare

MIMSunit 0.11.0

BREAKING CHANGES

  • Add a new input argument use_snapshot_to_check to mims_unit() to configure the data used to check duplicated timestamps.

  • Enable has_ts argument in import_actigraph_csv() and import_actigraph_csv_chunked() functions. Now the timestamps will be computed based on the sampling rate and start time in the actigraph csv header.

Bug fixes

  • Check duplicated timestamps and raise error when calling mims_unit(). Use diff function to accelerate the computation (@vincentvanhees).
  • Fix the rest_on_table sample data to remove duplicated timestamps in it.

Documentation

  • Update FAQ about the duplicated timestamps problem.

v0.10.0

29 Apr 18:25
Compare
Choose a tag to compare

BREAKING CHANGES

  • import_actigraph_csv() and import_actigraph_csv_chunked() functions no longer support has_ts flag. Users should always ensure input data have timestamps in the first column. Note that for versions <=0.9.2, always make sure the input raw accelerometer data includes timestamps in the first column and set has_ts to TRUE when using import_actigraph_csv() and import_actigraph_csv_chunked() functions to avoid a known bug.

Bug fixes

  • Remove warnings from dplyr > 1.0.0.
  • The first column of the input data.frame does not have to be HEADER_TIME_STAMP. #29.

Refactor

  • Refactor export_to_actilife() and conceptual_diagram vignette to be compatible with newer versions of readr and plyr packages.

Development

  • Update development package versions

Documentation

  • Add back man pages to the github development version.
  • Fix invalid urls in any documentations.

v0.9.1

20 May 23:18
Compare
Choose a tag to compare

Features

  1. Added a function shiny_app() to start a local shiny app that allows users to load local files to compute mims unit.
  2. values and provides an interactive graph to view the computed values.

Improvement

  1. Optimize memory usage when computing MIMS unit values.

Documentation

  1. Added datasets article vignette("datasets").

v0.9.0

26 Apr 14:44
Compare
Choose a tag to compare

MIMSunit 0.9.0

Bug fixes

  • Check column existence before coverting data type in import_mhealth_csv().
  • Quick failure if dynamic_range is not specified.
  • Auto set has_ts to False when there are only three columns in the input data in import_actigraph_csv().

Features

  • Sort timestamps of input data before computing mims unit values in custom_mims_unit().

Documentation

  • Indicate columns cannot be reordered in the argument description of extrapolate() and mims_unit().

CI

  • Add appveyor for non-Github actions testing

v0.8.2

02 Mar 17:04
Compare
Choose a tag to compare

MIMSunit 0.8.2

Bug fixes

  • Make sure codes do not change options.
  • Fix timestamps in import_actigraph_csv_chunked().

MIMSunit 0.8.1

New features

  • Add two new sample data edge_case and rest_on_table.
  • Add one new external sample data actigraph_timestamped.csv.

Improvement

  • Use tempfile in examples.

Documentation

  • Remove unused vignettes.
  • Add documentation for the new sample data.
  • Use internal data for all vignettes.

MIMSunit 0.8.0

New features

  • New import data functions import_actigraph_csv_chunked() and import_mhealth_csv_chunked() to support loading large csv files in chunks. See issue #11.

Improvement

  • Use chunked import functions in mims_unit_from_files() to support large input files.

Bug fixes

  • Column naming in vector_magnitude() and sum_up() functions.
  • Crashing bug in plotting functions.
  • Chances of incorrect parsing of time zone during interpolation and extrapolation.

Documentation

  • Add examples to functions.

Minor changes

  • show_progress is discarded in top-level mims_unit() functions. Use use_gui_progress instead.

BREAKING CHANGES

  • Minimal R version requirement is bumped to 3.6.0.

v0.7.0

22 Jan 20:42
Compare
Choose a tag to compare

Known issue

  • import_* functions may fail due to memory overflow if loading a large csv file on a low-memory computer. See #11

Improved

  • Refine vignettes codes and texts.
  • Better progress bar support for MIMSunit::mims_unit functions. Use show_progress to control whether to show progress bar during computation.

Bug fixes

  • Fix missing "testthat" dependency.
  • Add stringr namespace before str_detect.

Dependency changes

  • Remove dependency on "akima". This package is not used.

Breaking changes

  • Remove dependency on "mHealthR" package.
  • Add top-level API function MIMSunit::mims_unit_from_files.
  • Add new plotting function MIMSunit::generate_interactive_plot.