René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Consuming more time to pass object from .net to python ","articleBody":"### Environment\r\n\r\n-   Pythonnet version: master\r\n-   Python version: 3.5\r\n-   Operating System: Windows 7\r\n\r\n### Details\r\n\r\n-   I am calling python function from my c# code . Initially I import python module using Py.Import function . After importing , will pass a List\u003cfloat\u003e data type of size 299x299x3 to python function as argument . This passing of argument taking 1.5 seconds where as a block of  python code takes only 90 msec . \r\n\r\nHere is  c# code\r\n\r\n```csharp\r\nusing(Py.GIL())\r\n{\r\ndynamic Python_predict = Py.Import(\"Prediction\");\r\nPython_predict.__init__(modelpath);\r\n\r\n\r\nList\u003cfloat\u003e imagelist=GetNormalizedFloatBGRImage(image).Data.Cast\u003cfloat\u003e().ToList();\r\nDateTime t1 = System.DateTime.Now;\r\nPredicted_Type=Python_predict.Findtype(imagelist);\r\nDateTime t2 = System.DateTime.Now;\r\nConsole.WriteLine(\"Predict timing from c# {0} seconds\", (t2.Second + ((float)t2.Millisecond / 1000)) - (t1.Second + ((float)t1.Millisecond / 1000)));\r\n\r\n}\r\n```\r\n\r\nHere is  python code\r\n\r\n```python\r\ndef __init__(model_path):\r\n        kr.model = kr.load_model(model_path) #Model is loaded only once\r\ndef Findtype(ImageData):\r\n        t1=time.time()\r\n        N_array = np.array(ImageData,dtype=np.float64)#creates Numpy array from .net List\u003cfloat\u003e array\r\n        N_array_reshaped=np.reshape(N_array,(299,299,3))\r\n\r\n        #convert BGR to RGB\r\n        temp_array[:,:,0] = N_array_reshaped[:,:,2]#bgR -\u003e Rgb \r\n        temp_array[:,:,2] = N_array_reshaped[:,:,0]#Bgr -\u003e rgB\r\n        temp_array[:,:,1] = N_array_reshaped[:,:,1]#bGr -\u003e rGb\r\n        img_list[0]=temp_array\r\n        \r\n        #prediction = kr.model.predict(img_list, batch_size=1, verbose=0) # Predictions can be made throughout the lifecycle of MakePrediction class's instance\r\n        #predicted_class = np.argmax(prediction, axis=1)\r\n        t2=time.time()\r\n        print('time on python ',t2-t1)\r\n        return predicted_class[0] + 1\r\n ```       \r\n\r\nSo c# code send image array of size 299*299*3 of float time to python , by which it gets converted into numpy array and makes the prediction on gpu . \r\nif  i run the standalone python code , it takes 90msec but if i call from .net 1.5 sec for passing an argument . \r\n\r\nKindly help me to reduce the timing since it is for real time application .\r\n\r\nhere is a attached screenshots\r\n![untitled](https://user-images.githubusercontent.com/38282699/38613443-95ad4652-3da7-11e8-8934-855b31c48224.png)\r\n","author":{"url":"https://github.com/48naveen","@type":"Person","name":"48naveen"},"datePublished":"2018-04-11T11:17:54.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":10},"url":"https://github.com/658/pythonnet/issues/658"}