Chunk / ImageR

Download as
Browse Code ยป

Commit 5c9c44da22fd65b593e5b7b7cf5d788c5f11a930

Authored by Chunk
1 parent 54e2adda
Exists in master and in 1 other branch refactor

staged.

Showing 1 changed file with 29 additions and 27 deletions   Show diff stats
test/test_whole.py
  Diff comments   View file @5c9c44d
... ... @@ -4,6 +4,7 @@ from ..mspark import SC
4 4 from pyspark.mllib.regression import LabeledPoint
5 5 import happybase
6 6  
  7 +
7 8 def test_whole():
8 9 cols0 = [
9 10 'cf_pic:data',
... ... @@ -33,7 +34,7 @@ def test_whole():
33 34  
34 35 # rdd_data = sparker.read_hbase("ILSVRC2013_DET_val-Test_1", func=SC.rddparse_data_ILS, collect=False) \
35 36 # .mapValues(lambda data: [data] + SC.rddinfo_ILS(data)) \
36   - # .flatMap(lambda x: SC.rddembed_ILS_EXT(x, rate=0.2)) \
  37 + # .flatMap(lambda x: SC.rddembed_ILS_EXT(x, rate=0.2)) \
37 38 # .mapValues(lambda items: SC.rddfeat_ILS(items))
38 39  
39 40 rdd_data = sparker.read_hbase("ILSVRC2013_DET_val-Test_1", func=SC.rddparse_data_ILS, collect=False).mapValues(
... ... @@ -64,36 +65,37 @@ def test_whole_ext():
64 65 cols = ['cf_pic:data']
65 66 list_data = []
66 67 for key, data in table.scan(columns=cols):
67   - data = data['cf_pic:data']
68   - list_data.append((key,data))
  68 + data = data['cf_pic:data']
  69 + list_data.append((key, data))
69 70  
70 71 sparker = SC.Sparker(host='HPC-server', appname='ImageILSVRC-S', master='spark://HPC-server:7077')
71   - rdd_data = sparker.sc.parallelize(list_data,20)\
72   - .mapValues(lambda data: [data] + SC.rddinfo_ILS(data))\
73   - .flatMap(lambda x: SC.rddembed_ILS_EXT(x, rate=0.2))\
  72 + rdd_data = sparker.sc.parallelize(list_data, 40) \
  73 + .mapValues(lambda data: [data] + SC.rddinfo_ILS(data)) \
  74 + .flatMap(lambda x: SC.rddembed_ILS_EXT(x, rate=0.2)) \
74 75 .mapValues(lambda items: SC.rddfeat_ILS(items))
75 76  
76   - rrr = rdd_data.collect()
77   - print "-----------------",len(rrr),"===================="
78   - print "+++++++++++++++++",rrr[0],"**********************"
79   - # try:
80   - # with table.batch(batch_size=5000) as b:
81   - # for imgname, imginfo in rdd_data.collect().items():
82   - # b.put(imgname,
83   - # {
84   - # 'cf_pic:data': imginfo[0],
85   - # 'cf_info:width': str(imginfo[1]),
86   - # 'cf_info:height': str(imginfo[2]),
87   - # 'cf_info:size': str(imginfo[3]),
88   - # 'cf_info:capacity': str(imginfo[4]),
89   - # 'cf_info:quality': str(imginfo[5]),
90   - # 'cf_info:rate': str(imginfo[6]),
91   - # 'cf_tag:chosen': str(imginfo[7]),
92   - # 'cf_tag:class': str(imginfo[8]),
93   - # 'cf_feat:' + feattype: imginfo[9],
94   - # })
95   - # except ValueError:
96   - # raise
  77 + # rrr = rdd_data.collect()
  78 + # print "-----------------", len(rrr), "===================="
  79 + # print "+++++++++++++++++", rrr[0], "**********************"
  80 + try:
  81 + with table.batch(batch_size=5000) as b:
  82 + for item in rdd_data.collect():
  83 + imgname, imginfo = item[0], item[1]
  84 + b.put(imgname,
  85 + {
  86 + 'cf_pic:data': imginfo[0],
  87 + 'cf_info:width': str(imginfo[1]),
  88 + 'cf_info:height': str(imginfo[2]),
  89 + 'cf_info:size': str(imginfo[3]),
  90 + 'cf_info:capacity': str(imginfo[4]),
  91 + 'cf_info:quality': str(imginfo[5]),
  92 + 'cf_info:rate': str(imginfo[6]),
  93 + 'cf_tag:chosen': str(imginfo[7]),
  94 + 'cf_tag:class': str(imginfo[8]),
  95 + 'cf_feat:ibd' : imginfo[9],
  96 + })
  97 + except ValueError:
  98 + raise
97 99  
98 100  
99 101  
... ...