gistfile1.py

import numpy as np
import pandas as pd
from scipy import stats
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns

combined = pd.read_csv('/Users/daniel.lurie/Dropbox/CMI/INDI/CoRR/qc/temp_motion_326.csv', sep=',')
snr = pd.read_csv('/Users/daniel.lurie/Dropbox/CMI/INDI/CoRR/qc/tsnr_temp_328.csv', sep=',')

class Site():
	def __init__(self, site_name, sub_range):
		self.site_name = site_name
		self.sub_range = sub_range
		self.snr_df_name = self.site_name + "_snr"
		self.motion_data = None
		self.signal_data = None

		self.slice_site()

	# Define a function that creates site-specific data frames based on a range of subject numbers (specified as regex)
	# One data frame for motion data, one for SNR data
	# Prints number of rows in each new DF
	def slice_site(self):
		self.motion_data = combined[combined['Subject'].str.contains(self.sub_range)]
		self.signal_data = snr[snr['subject'].str.contains(self.sub_range)]
		print "Scans in motion data = "+str(len(self.motion_data.index))
		print "Scans in SNR data = "+str(len(self.signal_data.index))
		
	# for the love of god, rename me
	def make_pretty_thing(self):
		func_quad(site_name = self.site_name, meanfd_column = self.motion_data.MeanFD, 
          		meandvars_column = self.motion_data.MeanDVARS,
          		pctvols_column = self.motion_data[self.motion_data.columns[37]],
          		tsnr_column = self.signal_data.tsnr)

# Utah 26017-26043 (178)
utah = Site(site_name="Utah", sub_range="2601[7-9]|260[2-3][0-9]|2604[0-3]")
utah.make_pretty_thing()