Skip to content

Average

Jump to bottom Edit New page
André Mouraux edited this page Apr 10, 2015 · 3 revisions

Average epochs

Average epochs across trials.

average_epochs

  • Operation. In addition to computing the average, it is also possible to compute the median and the standard deviation across trials.

Sliding operation along dimension

Apply a sliding average (or other operation) along a given dimension (X, Y or Z).

sliding_operation

This function will perform a sliding operation along a given dimension. The width of the window is defined by the window width. For each sample of the dataset along the chosen dimension, the operation will be computed using all samples within the window, centered around that sample.

For example, if one selects to compute a sliding average along the X dimension with a window width of 0.2, the value of the output dataset at latency xi will correspond to the mean of the values obtained at xi-0.1 s and xi+0.1s

  • Operation. In addition to applying a sliding average along a given dimension, it is also possible to compute the standard deviation, maximum, minimum, 75th percentile and 25th percentile along a given dimension.
  • **Slide along dimension.__ The dimension along which the sliding operation should be performed.
  • Window width. The width of the window used to compute the sliding operation.

Sliding average across trials

Apply a sliding average across trials (also known as "ERP image"). This can be used to compute maps expressing signal amplitude as a function of time (x-axis) and trial (y-axis).

sliding_average

  • Number of lines. The number of lines (y-axis) of the output ERPimage.
  • Use entire epoch range. This will compute the ERPimage across the entire duration of the epochs.
  • X-axis start and X-axis end define the beginning and end of the X-axis of the output ERP image. In this example, the ERPimage will be computed using the data sampled between 0 and 1.
  • Smooth using a Hanning function. If checked, each line of the ERPimage will not correspond to the data measured at a single epoch, but to a weighted average of the data measured at that epoch and surrounding epoch. The weights are defined using a Hanning function. The width of the Hanning function is defined by the Hanning width field.

###Grand average (weighted)

Compute a grand average of the signals contained in multiple datasets. This function also makes it possible to compute a weighted grand average, such as to give different weights to each dataset (for example, to take into consideration that the averages obtained in each dataset were not obtained using the same number of trials).

grand_average

  • Datasets. The selected datasets.
  • Weight. The weight to assign to each dataset (default = 1).

###Linear channel map

Compute a linear channel map. This will generate a map interpolating the values obtained at the different channels along the y dimension. This is useful to explore signals obtained from a linear configuration of electrodes, for example, to assess the occurrence of polarity reversals along the different contacts of an intracerebral electrode tract. In most cases, you will want to apply this function to the signals obtained after applying a linear CSD.

linear_channel_map

  • Number of lines for channel interpolation. The number of lines to interpolate when creating the linear channel map.

###Pool channels

Pool (average) channels of a dataset.

pool_channels

  • Pool channels. The channel(s) to pool (average).
  • Mixed channel name. The name of the new (pooled) channel.

###Weighted channel average (create template)

Create a template to compute a weighted channel average. This is useful if you wish to express the magnitude of a given response measured at different electrodes.

This function is still EXPERIMENTAL.

weighted_channel_average_template

The function will search for the channels displaying maximum or minimum amplitude at a given position of the X/Y/Z axis. A template will then be created by selecting N channels exhibiting the strongest response. Each channel will be assigned a weight proportional to the amplitude of the response at the given position.

  • Search within channels. The channels within which channels exhibiting the strongest response magnitude should be searched.
  • Number of channels to include in the template. The number of channels to be included in the template.
  • X/Y/Z. The location of the response along the X/Y/Z dimension (e.g. the latency of a peak, or the latency and frequency of a response identified in the time-frequency domain).
  • Maximum/Minimum. Choose maximum if the response is positive (e.g. a positive peak, event-related synchronization). Choose minimum if the response is negative (e.g. a negative peak, event-related desynchronisation).
  • Normalize template. Normalize the weights such that their sum equals 1.

###Weighted channel average (apply template)

Apply a weighted channel average template to one or more datasets.

This function is still EXPERIMENTAL.

weighted_channel_average_apply

  • Load template. Load a dataset containing a weighted channel average template (computed using the weighted channel average (create template) function.
Add a custom footer

Plugins

User interface

File

Edit

Events

Pre-processing

Post-processing

Statistics

Figures

Clone this wiki locally