Revise langsamp01.md

[tashrifbillah]

cd /data/predict1/to_nda/nda-submissions/network_combined
/data/predict1/utility/nda-transform/image03.py --dict langsamp01 --shared ndar_subject01.csv --root /data/predict1/data_from_nda/ --template "Pr*/PHOENIX/GENERAL/*/processed/*/surveys/*.Pr*.json" --data ../langsamp01/merged_language_fluency_nda.csv -o langsamp01_derived.csv

(merged_language_fluency_nda.csv has visit column and it is appropriately populated)

[tashrifbillah]

Dheshan will provide redcap_event_name column. But _langsamp01.py will replace that with visit column.

./_langsamp01.sh open merged_language_fluency.csv
./_langsamp01.sh psychs merged_language_fluency.csv

(merged_language_fluency.csv has redcap_event_name column and it is appropriately populated)

[tashrifbillah]

Add the following lines to tutorial. Basically, we are teaching collaborators how to get REDCap features for a transcript:

utility/nda-transform/langsamp01_qc.py

Lines 42 to 67 in c70a01f

	ngroups=derived.groupby(['src_subject_id', 'interview_type','visit']).ngroups
	derived.set_index(['src_subject_id', 'interview_type','visit'], inplace=True)
	# sort_index() to avoid
	# PerformanceWarning: indexing past lexsort depth may impact performance.
	derived.sort_index(inplace=True)
	for i,f in enumerate(files):
	df=pd.read_csv(f,dtype=str,header=1)
	if i==0:
	df1=df.copy()
	else:
	df1=pd.concat((df1,df),ignore_index=True)
	print('\tabsents:')
	absent=[]
	for i,row in df1.iterrows():
	s=row['src_subject_id']
	e=row['visit']
	t=row['interview_type']
	try:
	derived.loc[(s,t,e)]
	except:
	absent.append(f" ['{s}','{t}','{e}'] ")