filterRecords.py

import synapseclient
import pandas as pd

syn=synapseclient.Synapse(skip_checks=True)
syn.login(silent=True) 


records = pd.read_csv('publicRecordIds', sep='\t')
records = records.query('studyId=="parkinson"')

allVoiceData = syn.tableQuery('SELECT recordId, healthCode FROM syn4590865').asDataFrame()

idx = ['Voice' in x for x in records.schemaKey]
voiceRecordsToKeep = records.ix[idx, 'recordId']

#Find common healthCodes
df = allVoiceData[allVoiceData.recordId.isin(voiceRecordsToKeep)]
df = df.groupby('healthCode')['healthCode'].count()
df.sort()
print '\',\''.join(df[-50:].index)