I want to use C++ to load TensorFlow model. And I want to know size of model's input, which is the placeholder in the model.
I google this problem, but I just find this link in stackoverflow :
C++ equivalent of python: tf.Graph.get_tensor_by_name() in Tensorflow?
Although I can get node, but tensorflow document don't tell me how to access the size of the node. So is there anyone know something about this?
Thank you so much!
OK，after many times attempts. I have find a workaround solution, It maybe tricky but works well.
At first, we can get the placeholder node using following code:
GraphDef mygd = graph_def.graph_def();
for (int i = 0; i < mygd.node_size(); i++)
{
if (mygd.node(i).name() == input_name)
{
auto node = mygd.node(i);
}
}
Then through the NodeDef.pd.h(tensorflow/core/framework/node_def.pb.h), we can get AttrValue through code like below:
auto attr = node.attr();
Then through the attr_value.cc(tensorflow/core/framework/attr_value.cc), we can get the shape attr value through code like below:
tensorflow::AttrValue shape = attr["shape"];
and the shape AttrValue is the structure used to store shape information. We can get the detail information through the function SummarizeAttrValue in tensorflow/core/framework/attr_value_util.h
string size_summary = SummarizeAttrValue(shape);
And then we can get the string format of shape like below:
[?,1024]
Related
Sorry for the newbie question. Learning Apex here. Been working on an issue for several hours and can't seem to needle it out.
I have a JSON that I need converted to a List... the JSON is retrieved via an API. The code is pretty simple, it is only a couple of lines.
The JSON looks like this:
{"id":1,"abbreviation":"ATL","city":"Atlanta","conference":"East","division":"Southeast","full_name":"Atlanta Hawks","name":"Hawks"}
And the code I am told to use to retrieve it looks like this:
Map<String, Object> resultsMap = (Map<String, Object>) JSON.deserializeUntyped(results.getBody());
Based on the JSON provided, there does not appear to be a MAP key being assigned, so I have no idea how to get it so that I may assign it to a List...
I've already tried assigning it directly to a List, but I didn't get much success there either.
I've tried this already:
List<Object> other = (List<Object>) results.get('');
I've also tried this:
List<Object> other = (List<Object>) results.keySet()[0];
I'm sure it is something simple. Any help would be appreciated.
Thanks
You cant convert the map to a list without an id. Alternatively why to convert it to List. Use the map itself. Below is a code example of how you can use it using your JSON response. For example, I want to insert this JSON response into the contact record, so I can map it using the MAP itself.:
#RestResource(urlMapping='/URIId/*')
global class OwneriCRM {
#HttpPut
global static String UpdateOwnerinfo(){
RestRequest req= RestContext.request;
RestResponse res= RestContext.response;
res.addHeader('Content-type','application/JSON');
Map<String,Object> responseresult=new Map<String,Object>();
Map<String,Object> results= (Map<String,Object>)JSON.deserializeUntyped(req.requestBody.toString());
List<contact> insertList = new List<contact>();
for (String key : results.keySet()){
System.debug(results.get(key));
contact c= new contact();
c.ContactField__C = results.get(id);
c.ContactField1__C = results.get(city);
c.ContactField2__C = results.get(conference);
c.ContactField3__C = results.get(division);
c.ContactField3__C = results.get(full_name);
c.ContactField3__C = results.get(name);
insertList.add(c);
}
if(!insertList.Isempty()){
update insertList;
}
Question can be disregarded. Not sure if there is a way to withdraw the question.
Question was based on the fact that previous APIs had been sent in the following format:
{"data": {"more stuff": "stuff"} }
And the map key was 'data' with a list. In this scenario, the whole API was a list and the keys to the map were set in place with the actual values instead.
I am new to Python and am trying to get my ahead around parsing SSE client code. I am using the SSE Client library. My code is very basic and follows the sample exactly. Here it is:
from sseclient import SSEClient
devID = "xxx"
AToken = "xxx"
sparkURL = 'https://api.spark.io/v1/devices/' + devID + '/events/?access_token=' + AToken
messages = SSEClient(sparkURL)
for msg in messages:
print(msg)
print(type(msg))
The code runs without a problem and I see some blank lines and SSE data coming through. Here is the sample output:
<class 'sseclient.Event'>
{"data":"0 days, 0:54:43","ttl":"60","published_at":"2015-04-09T22:43:52.084Z","coreid":"xxxx"}
<class 'sseclient.Event'>
<class 'sseclient.Event'>
{"data":"0 days, 0:55:3","ttl":"60","published_at":"2015-04-09T22:44:12.092Z","coreid":"xxx"}
<class 'sseclient.Event'>
The actual output above looks like a dictionary, but its type is "sseclient.Event". I am trying to figure out how to parse the output so I can pull out one of the fields and nothing I have tried has worked.
Sorry if this is basic questions, but can someone provide some simple guidance on how I would either convert the entire output to a dictionary or perhaps just pull out one of the fields?
Thank you in advance!
I figured this out. In case anyone else experiences the same problem, here is how I got it to work. The key was using msg.data and not just msg. I then converted the out using the JSON library and am good to go.
messages = SSEClient(sparkURL)
for msg in messages:
outputMsg = msg.data
if type(outputMsg) is not str:
outputJS = json.loads(outputMsg)
FilterName = "data"
#print( FilterName, outputJS[FilterName] )
print(outputJS[FilterName])
I'm trying to control Word through a c++ builder 5 application. I would like to
open a ".dot" model file created with Word and modify it. In the ".dot" model file
there are some fields. For example, Title, LastName, FirstName, Address
and so on, and I would like to modify these fields putting text into them and then
saving file with a new name, for example "Warning.doc" leaving the ".dot" file
unaltered.
I can open the file, count the number of fields it contains, but then
when it comes to replacing each field with a string I don't know how to do because
I don't have a complete documentation on OleFunction and OlePropertyGet methods. I attach my source code to this message, can anybody help me to solve this problem please?
try
{
my_word = Variant::CreateObject("word.application");
}
catch (...)
{
Application->MessageBox("Unable to obtain Word automation object",
"Error:",MB_OK | MB_ICONERROR);
}
my_word.OlePropertySet("Visible", (Variant)true);
void __fastcall TForm1::Button2Click(TObject *Sender)
{
Variant this_doc;
Variant my_fields;
Variant test;
int k,field_count;
AnsiString test1;
AnsiString filename = "d:\\ProgrammaWord\\1-Avviso.dot";
my_docs = my_word.OlePropertyGet("Documents");
this_doc = my_docs.OleFunction("Open", filename);
my_fields = this_doc.OlePropertyGet("Fields");
field_count = my_fields.OlePropertyGet("Count");
for(k = 1; k <= field_count; k++)
{
test = my_fields.OleFunction("Item",(Variant)k);
test1 = test.OleFunction("Value"); //This instruction throws an exception
// "Value" is not a recognized parameter
// in this case
Memo1->Lines->Add(test1);
}
}
I never used word Ole but I used it for Outlook and Excel, I can't try it with word since I'm currently on OSX but you should try something similar as what I did.
The generic way of using Ole was to OleGetproperty() while you get the targeted field then OleSetProperty("action", ...).
for example when I wanted to change the color of the text in a particular cell of my excel document I used:
Variant _excel = Variant::CreateObject("Excel.Application");
Variant _workbook = _excel.OlePropertyGet("WorkBooks").OleFunction("Open", filename);
Variant _worksheet = _workbook.OlePropertyGet("WorkSheets", sheet);
_worksheet.OlePropertyGet("Cells", row, col).OlePropertyGet("Font").OlePropertySet("Color", color);
Here I instanciate an excel object, then I load a file into it (_workbook), then I select the _worksheet from the _workbook and I start my business.
Here comes the interesting part:
It concist of getting to a particular cell, getting the font object from it, and then setting the color of this font object.
Disclaimer: This is an example from my sources for excel, it's not directly related to your example, but maybe you can understand the principe with it. I can't try to figure out what you need because I have no windows right now.
Hope this can help you. Finding ressources for OLE can be fierce if you don't have the good patterns to look for.
I see this sample:
http://www.aspose.com/docs/display/slidesnet/Creating+a+Table+from+Scratch
...and when I try to use it, I see that the library has changed (I guess) and it needs me to do sth like that:
int lIndex = pSld.Shapes.AddTable(pUppLeftPoint.X, pUppLeftPoint.Y, pColumnWidths, pRowWidths);
TableEx lTable = (TableEx)pSld.Shapes[lIndex];
(I can only cast to TableEx and not to Table)
But I cannot find how to get the cell's TextFrame. The site says :
TextFrame tf = table.GetCell(0, 0).TextFrame;
But I have nothing like this...
Am I missing sth?
Any ideas?
EDIT :
I found out that my code is for PPTX and the site's code for PPT:
http://www.aspose.com/community/forums/thread/453613/facing-performance-issue-due-to-addtable-method-of-pptx.aspx
But, again, how do you get the cell's contents in PPTX?
Ha! I got it!
There's an indexer:
lTable[i,j]
I'm having a problem with my picker in one of my apps. I have an NSDictionary obtained from a property list that contains a bunch of keys, which in turn contain a bunch of strings. I have two components, each one should have the same list of strings within. I also have a slider that I want to use to allow the user to change keys. So when the slider's value goes from 0 to 1 the key at index 1 in the dictionary should load its contents into the pickerview's components.
It's working as far as loading the new contents into the picker based on the slider. I've been using the slider's tag as the variable to dictate which contents get loaded. The problem is that after loading a new list of items the program crashes, I'm thinking that the number of rows needed isn't getting update or something but I'm just not experienced enough with UIPickerView to isolate the problem myself without spending more hours than I've already used trying to figure this out myself.
Here are my delegate/data methods for the picker:
#pragma mark -
#pragma mark Picker Delegate/Data Methods
-(NSInteger)numberOfComponentsInPickerView:(UIPickerView *)pickerView
{
return 2;
}
-(NSInteger)pickerView:(UIPickerView *)pickerView numberOfRowsInComponent:(NSInteger)component
{
//aryNames is an NSArray instance variable that contains the values for the components of the picker
if (component == 0)
return [self.aryNames count];
return [self.aryNames count];
}
-(NSString *)pickerView:(UIPickerView *)pickerView titleForRow:(NSInteger)row
forComponent:(NSInteger)component
{
//I think this is where my problem is
//I'm using a string to select the object
// at the index of the slider's location to
// fill up the instance variable with new data.
//Anyway, it works fine if I have two different arrays hardcoded
//but I'd really like to have this load dynamically because
//there are a lot of keys and this way I could add and remove keys without
//worrying about changing code
NSString *selectedType = [self.aryKeys objectAtIndex:slideUnitTypes.tag];
NSArray *newNames = [dictAllNames objectForKey:selectedType];
self.aryNames = newNames;
return [aryNames objectAtIndex:row];
}
//I'm pretty sure that the method below is not the problem
-(void)pickerView:(UIPickerView *)pickerView didSelectRow:(NSInteger)row inComponent:
(NSInteger)component
{
if (component == 0)
{
[firstValueHeading setText:[aryNames objectAtIndex:row]];
}
else
{
[secondValueHeading setText:[aryNames objectAtIndex:row]];
}
}
If it wasn't descriptive enough or you need to see more of my code please tell me. This problem has been a real bugger in an otherwise smooth project. Thanks.
I am still fairly new to this myself, but in Troy Brant's book (chapter 9) he does this. You should grab the book from the library/bookstore and review the source code at http://troybrant.net/iphonebook/chapter9/Ruralfork-done.zip
It should help.
i've actually solved this long since. here is the code for that delegate if it helps:
-(NSString *)pickerView:(UIPickerView *)pickerView titleForRow:(NSInteger)row forComponent:(NSInteger)component
{
NSUInteger index = slideUnitTypes.value;
NSString *placeString = [self.aryKeys objectAtIndex:index];
NSArray *returnThisArray = [dictAllNames objectForKey:placeString];
return [returnThisArray objectAtIndex:row];
}
if anyone out there needs to see any of my other delegates just comment on this answer and hopefully SO should send me an email.