Python toolbox with general-purpose functions for behavioral and neuroimaging research
pip install git+https://github.com/LabSocialAffectNeuro/lsan_tools
- bids==0.9.5
- pandas>=0.24
- glob
- nipype==1.2.3
- numpy>=1.16.4
- os
- sklearn==0.21.2
Python class used to perform manipulations on self-report surveys (e.g., score questionnaires, extract specific participants' data, etc.)
Scoring functions:
survey.score_hexaco(): Scores HEXACO Personality Inventory (60-item)survey.score_rel_mobility(): Scores Relational Mobility Scale (12-item)survey.score_isel(): Scores Interpersonal Support Evaluation List (40-item)survey.score_dospert(): Scores Domain-Specific Risk-Taking (DOSPERT) Scale (60-item)
Other functions:
lsan_survey.select_data("sub_ids.txt", rewrite_to_self=True): Selects specific subject data from survey using sub_ids.txt file, rewrites survey.data in class, but does not save as comma-separated filesurvey.retain_items(list): Retains specific question items (e.g., demographics)survey.join_data(): Joins scored surveys and save as comma-delimited file
from lsan_tools.behav import survey
survey = survey('survey.csv', 'PIN') # load data into lsan_survey class
survey.select_data([101,102], rewrite_to_self=True, save=False) # only select subject IDs 101 and 102 from original data
survey.score_hexaco() # score subscales in HEXACO Personality Inventory (60-item)
survey.score_isel() # score subscales in the Interpersonal Support Evaluation List (40-item)
survey.retain_items(['age','gender']) # retain reports about single question items such as age or gender
# join all scored data and store in new DataFrame called new_survey and save to .csv called "scored_survey_output.csv"
survey.join_data(filename="scored_survey_output")
(in development) Python class used to take BIDS-formatted events.tsv files in a base BIDS directory and convert them for first-level GLM analysis in AFNI and SPM (FSL pending).
Functions:
events_class.get_timing(): Throws timing information for each subject into dictionary and writes event timing files to a BIDS derivatives folder for first-level GLM analysis in AFNI or SPM (Default: AFNI); Output contains a .txt file for each condition within a 'trial_type' column with N lines corresponding to N runs; each line is tab-separated and written with the heuristic ONSET_TIME:DURATION
Functions (in-development):
events_class.convert_secs_to_TRs(): Converts event timing information from secs to TRs for Design Matrix generation (in-development)events_class.gen_DesignMat(): Convolves specified events with specified HRF, adds nuissance regressors, and creates a Design Matrix for use in first-level GLM analysis (in-development)events.bunch_timing(): Throws timing information for each subject into dictionary for first-level GLM analysis in nipype (in-development)
from lsan_tools.fmri.postprep import events_class
base_dir = '/mnt/data/bids_dir'
task_id = 'task_name_specified_in_bids' #e.g., f'/mnt/data/bids_dir/sub-XX/func/sub-XX_run-XX_task-{task_id}_events.tsv'
events = events_class(base_dir, task_id)
# gets all trial types under task_id *events.tsv file
# (assumes first two columns: 'onset' and 'duration'),
# trims the onset times by 10 secs, and writes .txt files
# for AFNI first-level analysis to a BIDS derivatives folder called
# f'{bids_dir}/derivatives/timing/sub-XX/trial_type_column_name/*.txt'
events.get_timing(trimTRby=10, software='AFNI', write=True)
Some helpful math functions:
shuffle(df, type="pandas"): Takes a pandas DataFrame where the columns are variables and the observations are the rows (e.g., subject IDs), and randomly shuffles the row indicespairwise_vector(vec1, vec2, method="correlation", shuffle=False): Takes a pandas DataFrame and computes pairwise distance between two column vectorsstandardize(df, var_list): Takes pandas DataFrame and z-scores values within each of the columns
Some helpful data manipulation functions:
add_file_prefix(files_dir, prefix): Adds prefix to every file in a specified directory (wildcards infiles_dirstring work)