docstring bug: data consistency check (count_trials)

docstring states that data consistency checks are performed for stimuli, accuracy and confidence:

Lines 62 to 63 in ecfa900

    
               All trials where `stimuli` is not 0 or 1, accuracy is not 0 or 1, or confidence is 
        
               not in the range [1, nRatings], are automatically omitted.

in the implementation: stimuli, responses and confidence are checked.

cpm/cpm/utils/metad.py

Lines 122 to 131 in ecfa900

    
           # Check data consistency 
        
           tempstim, tempresp, tempratg = [], [], [] 
        
           for s, rp, rt in zip(stimuli, responses, confidence): 
        
               if (s == 0 or s == 1) and (rp == 0 or rp == 1) and (rt >= 1 and rt <= nRatings): 
        
                   tempstim.append(s) 
        
                   tempresp.append(rp) 
        
                   tempratg.append(rt) 
        
           stimuli = tempstim 
        
           responses = tempresp 
        
           confidence = tempratg

docstring should be changed to:
All trials where stimuli is not 0 or 1, responsesis not 0 or 1, or confidence is not in the range [1, nRatings], are automatically omitted.

Thanks for the report! The docstring was corrected as part of #106 (merged in #108, released in 0.26.1). The Notes section of count_trials now reads:

All trials where stimuli is not 0 or 1, the response is not 0 or 1, or confidence is not a whole number in the range [1, nRatings], are automatically omitted. When accuracy is provided, trials where accuracy is not 0 or 1 (including missing values) are omitted as well.

This matches the filtering in the implementation: responses (given or inferred from accuracy) are checked, and accuracy is checked as well when it is provided. A follow-up on branch 107-docstring-bug-data-consistency-check-count_trials also makes the "no valid trials" error name the same rules.

fix(metad): correct bin_ratings docstring, name every rule in count_t…

Merge pull request #109 from DevComPsy/107-docstring-bug-data-consist…

Navigation Menu

Repository navigation

docstring bug: data consistency check (count_trials) #107

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

	All trials where `stimuli` is not 0 or 1, accuracy is not 0 or 1, or confidence is
	not in the range [1, nRatings], are automatically omitted.

	# Check data consistency
	tempstim, tempresp, tempratg = [], [], []
	for s, rp, rt in zip(stimuli, responses, confidence):
	if (s == 0 or s == 1) and (rp == 0 or rp == 1) and (rt >= 1 and rt <= nRatings):
	tempstim.append(s)
	tempresp.append(rp)
	tempratg.append(rt)
	stimuli = tempstim
	responses = tempresp
	confidence = tempratg

docstring bug: data consistency check (count_trials) #107

Description

Activity

lenarddome commented on Nov 11, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions