-
Notifications
You must be signed in to change notification settings - Fork 44
/
GraphSelectTable.lua
executable file
·63 lines (53 loc) · 1.65 KB
/
GraphSelectTable.lua
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
local GraphSelectTable, parent = torch.class('nn.GraphSelectTable', 'nn.Module')
function GraphSelectTable:__init(index)
parent.__init(self)
self.index = index
--self.gradInput = {}
end
function GraphSelectTable:updateOutput(input)
-- handle negative indices
local index = self.index < 0 and #input + self.index + 1 or self.index
assert(input[index], "index does not exist in the input table")
self.output = input[index]
return self.output
end
local function zeroTableCopy(t1, t2)
for k, v in pairs(t2) do
if (torch.type(v) == "table") then
t1[k] = zeroTableCopy(t1[k] or {}, t2[k])
else
if not t1[k] then
t1[k] = v:clone():zero()
else
t1[k]:resizeAs(v)
t1[k]:zero()
end
end
end
for k, v in pairs(t1) do
if not t2[k] then
t1[k] = nil
end
end
return t1
end
function GraphSelectTable:updateGradInput(input, gradOutput)
-- make gradInput a zeroed copy of input
--zeroTableCopy(self.gradInput, input)
-- handle negative indices
local index = self.index < 0 and #input + self.index + 1 or self.index
-- copy into gradInput[index] (necessary for variable sized inputs)
--assert(self.gradInput[index])
--nn.utils.recursiveCopy(self.gradInput[index], gradOutput)
self.gradInput = gradOutput
return self.gradInput
end
function GraphSelectTable:type(type, tensorCache)
--self.gradInput = {}
self.output = {}
return parent.type(self, type, tensorCache)
end
function GraphSelectTable:__tostring__()
return torch.type(self) .. '(' .. self.index .. ')'
end
GraphSelectTable.clearState = nn.Identity.clearState