Skip to content

Latest commit

 

History

History
64 lines (44 loc) · 2.83 KB

NEWS.md

File metadata and controls

64 lines (44 loc) · 2.83 KB
Raw

NeuroDecodeR 0.2.0

New features

  • Added functionality to the the rm_main_results so it now calculates approximate standard deviations and standard errors for all decoding results. These decoding standard deviations/standard errors are likely a slight underestimate that one would get from new trials from the same set of neurons since they do not take into account that there is some dependencies between the test data when cross-validation is used. Also, added options to the rm_main_results plot() and plot_main_results()functions so that they can plot the standard deviations/standard errors as shaded regions.

Minor improvements and fixes

  • Changed the name of the private function get_center_bin_time() to be get_time_bin_center() (and deprecated the get_center_bin_time() function). Also added functions get_time_bin_start() and get_time_bin_end() to get the start and end times of time bin strings (these are not used by the package but could be useful for users of the package).

  • Changed read_matlab_raster_data() function so that if there is a field called trial_number in the MATLAB site_info, this this will automatically be set to be the trial_number in the R raster_data unless it is overwritten by setting sequential trial numbers using the add_sequential_trial_numbers argument.

  • Fixed the plot_main_results() function so that if 1 ms bins are used, times are plotting correctly rather than there being multiple points at a given time (which creating a vertical line artifact). This was done by rounding the times to be plotted using the floor() function rather than round() function.

  • Fixed the rm_confusion_matrix so that if a classifier does not return decision values, the and create_decision_vals_confusion_matrix is set to TRUE, a warning message will be given and create_decision_vals_confusion_matrix will be set to FALSE.

  • Added a warning message to the cv_standard constructor. If the result_metric argument is NULL, it will print a warning to let the user know that the default rm_main_rsults and rm_confusion_matrix result metrics are being used.

  • Updated log_load_results_from_result_name so that if multiple names match it returns a named list where the names are given by the manifest result_name column.

  • Updated plot_main_results so that if the results_to_plot is character vector and multiple manifest result_names match in the manifest file, it will plot all the results that match rather than give an error.

  • Small fix to the plot() function for the rm_main_results object so that it works with version 3.4.0 of ggplot2.

  • Fixed the package so that it works with tidyselect 1.2.0

NeuroDecodeR 0.1.0

Initial release!

  • This is the initial release of the NeuroDecodeR package. When new versions of the package are created, changes will be noted in this file.