TypeError: 'numpy.float64' object cannot be interpreted as an index

[jpsml]

This error is ocurring when I run main_train_dagnet for the first time. Do you have any clue of how to solve it?

Traceback (most recent call last): File "main_train_dagnet.py", line 1520, in linemod_train_realandsynth_main(cfgparser) File "main_train_dagnet.py", line 1378, in linemod_train_realandsynth_main visNetworkFiltersMontage(descrNet,resPath+startTimePidStr+"_filters_init.png") File "E:\ObjRecPoseEst-master\src\util\vis.py", line 264, in visNetworkFiltersMontage visFilterMontageOfLayer(wvals,fileName=layerFN,windowName=windowName) File "E:\ObjRecPoseEst-master\src\util\vis.py", line 217, in visFilterMontageOfLayer wAll = numpy.zeros((numRowsrowHeight,numColscolWidth),dtype=numpy.float32) TypeError: 'numpy.float64' object cannot be interpreted as an index

[Nemokang]

It seems that I am facing the same problems as you have experienced.Everything begins with the error in "fractNoise = cv2.medianBlur(fractNoise,3)" .Maybe errors come from our version of theano......

[Nemokang]

I have changed my version of numpy ,scipy and theano and fix errors successfully. My current version : NumPy 1.11 SciPy 0.16 theano 0.8