single-site fix

hase.py

@@ -102,6 +102,7 @@
 	parser.add_argument('-effect_intercept', action='store_true', default=False, help='Flag for add study effect to PD regression model')
 	parser.add_argument('-permute_ph', action='store_true', default=False, help='Flag for phenotype permutation')
 	parser.add_argument('-vcf', action='store_true', default=False, help='Flag for VCF data to convert')
+	parser.add_argument('-encoded', action='store_true', default=False, help='Flag to notify HASE that the input data are already encoded')
 	#TODO (low) save genotype after MAF
 	###
 
@@ -227,7 +228,7 @@
 		mapper.genotype_names=args.study_name
 		mapper.chunk_size=MAPPER_CHUNK_SIZE
 		mapper.reference_name=args.ref_name
-		mapper.load_flip(args.mapper)
+		mapper.load_flip(args.mapper, erase=args.encoded)
 		mapper.load(args.mapper)
 
 

hdgwas/data.py

@@ -172,11 +172,9 @@ def get_data(self,key,index):
 	def get_chunk(self,indices):
 		indices=self.link(indices)
 		indices=np.array(indices)
-		#print indices.shape
 		keys, ind = np.unique(indices[:,0], return_inverse=True)
 		result=None
 		gc.collect()
-		#print len(keys)
 		for i,k in enumerate(keys):
 			r=None
 			if k in self.loaded:
@@ -672,7 +670,10 @@ def __init__(self,path):
 			raise (e)
 		self.data_info={}
 		for i in self.files:
-			df=pd.read_csv(os.path.join(self.path,i), sep='\t', index_col=None)
+			for j in ['\t', ' ']:
+					df=pd.read_csv(os.path.join(self.path,i), sep=j, index_col=None)
+					if df.shape[1] > 1:
+						break
 			self.data_info[i]=np.array(df.columns[1:])
 
 

hdgwas/pard.py

@@ -53,19 +53,19 @@ def partial_derivatives(save_path=None,COV=None,PHEN=None, GEN=None,
 
     print ('Time to PD phenotype {} is {} s'.format(np.array(C).shape, t_phen.secs))
 
+    N_snps_read=0
     while True:
-
         with Timer() as t_gen:
-            SNPs_index, keys=MAP.get()
-            if isinstance(SNPs_index, type(None)):
+            genotype=GEN.get_next()
+            if isinstance(genotype, type(None)):
                 np.save(os.path.join(save_path,study_name+'_a_test.npy'), np.concatenate(a_test) )
                 np.save(os.path.join(save_path,study_name+'_metadata.npy'),metadata)
                 if B4_flag:
                     b4=np.concatenate(b4, axis=0)
                     np.save(os.path.join(save_path,study_name+'_b4.npy'),b4)
                 break
-            genotype=GEN.get(SNPs_index[0])
-            flip=MAP.flip[SNPs_index[0],0]
+            flip = MAP.flip[N_snps_read:N_snps_read + genotype.shape[0], 0]
+            N_snps_read += genotype.shape[0]
             flip_index=(flip==-1)
             genotype=np.apply_along_axis(lambda x: flip*(x-2*flip_index) ,0,genotype)
             genotype=genotype[:,row_index[0]]

hdgwas/tools.py

@@ -81,7 +81,7 @@ def summary(self):
 			print ('DF is not defined. Forced to use z_score statistics!')
 
 		if self.result_folder is None:
-			self.result_folder=glob.glob( os.path.join(self.result_path, '*.npy') )
+			self.result_folder=[ i for i in glob.glob( os.path.join(self.result_path, '*.npy') ) if 'RSID' not in i]
 
 		self.results['RSID']=np.array([])
 		self.results['p_value']=np.array([])
@@ -399,7 +399,7 @@ def push(self, new_keys,name=None, new_id=True):#TODO (middle) remove
 					self.dic[j]=[-1]*(self.n_study+1)
 					self.dic[j][self.n_study]=i
 
-	def load_flip(self,folder):
+	def load_flip(self,folder,erase=False):
 
 		if folder is None:
 			raise ValueError('Mapper is not defined!')
@@ -420,6 +420,8 @@ def load_flip(self,folder):
 				if len_flip!=self.flip[j].shape[0]:
 					raise ValueError('Different length of flip array between studies; used different ref panel!')
 		self.flip=np.array(self.flip).T
+		if erase:
+			self.flip.fill(1)
 
 	def load (self, folder):
 		if folder is None: