rm deprecated code

samplers.py

@@ -10,33 +10,6 @@ import torch
 from torch.utils.data.sampler import Sampler
 
 
-""" Deprecated
-class RandomIdentitySampler(Sampler):
-    def __init__(self, data_source, num_instances=4):
-        self.data_source = data_source
-        self.num_instances = num_instances
-        self.index_dic = defaultdict(list)
-        for index, (_, pid, _) in enumerate(data_source):
-            self.index_dic[pid].append(index)
-        self.pids = list(self.index_dic.keys())
-        self.num_identities = len(self.pids)
-
-    def __iter__(self):
-        indices = torch.randperm(self.num_identities)
-        ret = []
-        for i in indices:
-            pid = self.pids[i]
-            t = self.index_dic[pid]
-            replace = False if len(t) >= self.num_instances else True
-            t = np.random.choice(t, size=self.num_instances, replace=replace)
-            ret.extend(t)
-        return iter(ret)
-
-    def __len__(self):
-        return self.num_identities * self.num_instances
-"""
-
-
 class RandomIdentitySampler(Sampler):
     """
     Randomly sample N identities, then for each identity,