def fit_sx(mname, project='behav', regress=False):
pth=find_path()
data=defmod.find_data(mname, project)
grp_dict={}; subj_params=[]; aic_list=[]; bic_list=[]; dic_list=[]; ic_dict={}
for subj_idx, subj_data in data.groupby('subj_idx'):
m_sx=defmod.define_sxbayes(mname, subj_data, project=project, regress=regress)
m_sx.sample(1000, burn=500, dbname=str(subj_idx)+"_"+mname+'_traces.db', db='pickle')
sx_df=parse.stats_df(m_sx)
sx_df=sx_df.drop("sub", axis=1)
sx_df['sub']=[subj_idx]*len(sx_df)
subj_params.append(sx_df)
aic_list.append(aic(m_sx)); bic_list.append(bic(m_sx)); dic_list.append(m_sx.dic)
allsx_df=pd.concat(subj_params)
allsx_df.to_csv(mname+"_SxStats.csv", index=False)
ic_dict={'aic':aic_list, 'bic':bic_list, 'dic':dic_list}
ic_df=pd.DataFrame(ic_dict)
ic_df.to_csv(mname+"_IC_Rank.csv")
return allsx_df
def load_traces(m, mname, project='imaging'): pth=find_path() if project=='behav': m.load_db(pth+"beh_hddm/"+mname+"/"+mname+"_traces.db", db='pickle') else: m.load_db(pth+"img_hddm/"+mname+"/"+mname+"_traces.db", db='pickle') return m
def find_data(mname, project='imaging'): pth=find_path() if project=='behav': data=pd.read_csv(pth+"beh_hddm/allsx_feat.csv") else: data=pd.read_csv(pth+"img_hddm/allsx_ewma.csv") return data
def find_data(mname, project='imaging'):
pth = find_path()
if project == 'behav':
data = pd.read_csv(pth + "beh_hddm/allsx_feat.csv")
else:
data = pd.read_csv(pth + "img_hddm/allsx_ewma.csv")
return data
def load_traces(m, mname, project='imaging'):
pth = find_path()
if project == 'behav':
m.load_db(pth + "beh_hddm/" + mname + "/" + mname + "_traces.db",
db='pickle')
else:
m.load_db(pth + "img_hddm/" + mname + "/" + mname + "_traces.db",
db='pickle')
return m
def fit_sx(mname, project='behav', regress=False):
pth = find_path()
data = defmod.find_data(mname, project)
grp_dict = {}
subj_params = []
aic_list = []
bic_list = []
dic_list = []
ic_dict = {}
for subj_idx, subj_data in data.groupby('subj_idx'):
m_sx = defmod.define_sxbayes(mname,
subj_data,
project=project,
regress=regress)
m_sx.sample(1000,
burn=500,
dbname=str(subj_idx) + "_" + mname + '_traces.db',
db='pickle')
sx_df = parse.stats_df(m_sx)
sx_df = sx_df.drop("sub", axis=1)
sx_df['sub'] = [subj_idx] * len(sx_df)
subj_params.append(sx_df)
aic_list.append(aic(m_sx))
bic_list.append(bic(m_sx))
dic_list.append(m_sx.dic)
allsx_df = pd.concat(subj_params)
allsx_df.to_csv(mname + "_SxStats.csv", index=False)
ic_dict = {'aic': aic_list, 'bic': bic_list, 'dic': dic_list}
ic_df = pd.DataFrame(ic_dict)
ic_df.to_csv(mname + "_IC_Rank.csv")
return allsx_df
#!/usr/bin/env python
from __future__ import division
import hddm
import pandas as pd
import numpy as np
from patsy import dmatrix
import os
from mydata.munge import find_path
pth=find_path()
data=pd.read_csv(pth+"/beh_hddm/allsx_feat.csv")
def z_link_func(x, data=data):
stim = (np.asarray(dmatrix('0 + C(s,[[1],[-1]])', {'s':data.stimulus.ix[x.index]})))
return 1 / (1 + np.exp(-(x * stim)))
def v_link_func(x, data=data):
stim = (np.asarray(dmatrix('0 + C(s,[[1],[-1]])', {'s':data.stimulus.ix[x.index]})))
return x * stim
def define_model(mname, project='imaging', regress=False):
check_model(mname)
data=find_data(mname, project)
if project=='imaging':
intercept="b50N"
else:
#!/usr/bin/env python
from __future__ import division
import hddm
import pandas as pd
import numpy as np
from patsy import dmatrix
import os
from mydata.munge import find_path
pth = find_path()
data = pd.read_csv(pth + "/beh_hddm/allsx_feat.csv")
def z_link_func(x, data=data):
stim = (np.asarray(
dmatrix('0 + C(s,[[1],[-1]])', {'s': data.stimulus.ix[x.index]})))
return 1 / (1 + np.exp(-(x * stim)))
def v_link_func(x, data=data):
stim = (np.asarray(
dmatrix('0 + C(s,[[1],[-1]])', {'s': data.stimulus.ix[x.index]})))
return x * stim
def define_model(mname, project='imaging', regress=False):
check_model(mname)
data = find_data(mname, project)
