Shortcuts for common queries

Since most data other than EEG data can be naturally represented by a pandas DataFrame, querying and selecting subsets of data is generally very easy. Nevertheless, common queries can be tedious to have to write out every time, so shortcuts exist in the form of custom pandas accessors.

For example, we can get all word events for some events either by directly masking an events DataFrame or by using the shortcut accessor .events.words:

In [1]: from cmlreaders import CMLReader

In [2]: subjects = ["R1111M", "R1286J"]

In [3]: experiments = ["FR1"]

In [4]: all_events = CMLReader.load_events(subjects, experiments)

In [5]: all_events[all_events["type"] == "WORD"][:2]
Out[5]:
    answer                    eegfile  eegoffset  ...  subject       test  type
27    -999  R1111M_FR1_0_22Jan16_1638     100520  ...   R1111M  [0, 0, 0]  WORD
28    -999  R1111M_FR1_0_22Jan16_1638     101829  ...   R1111M  [0, 0, 0]  WORD

[2 rows x 24 columns]

In [6]: all_events.events.words[:2]
Out[6]:
    answer                    eegfile  eegoffset  ...  subject       test  type
27    -999  R1111M_FR1_0_22Jan16_1638     100520  ...   R1111M  [0, 0, 0]  WORD
28    -999  R1111M_FR1_0_22Jan16_1638     101829  ...   R1111M  [0, 0, 0]  WORD

[2 rows x 24 columns]

Available accessors

Upon importing cmlreaders, the following accessors are automatically registered with pandas.

class cmlreaders._accessors.events.EventsAccessor(obj)

Shortcuts for selecting events of various types.

Examples

>>> from cmlreaders import CMLReader
>>> reader = CMLReader("R1111M", "FR1", 0)
>>> df = reader.load("events")
>>> word_events = df.events.words
>>> stim_events = df.events.stim
>>> recalled_words = df.events.words_recalled
>>> forgotten_words = df.events.words_not_recalled

stim

Select all stim events.

stim_params

Expand the stim_params field in a friendly manner

words

Select all WORD onset events.

words_not_recalled

Select word events where the word was not recalled.

words_recalled

Select all recalled word events.