...

Loader = os.path.join(os.sep, r'C:\Users\...\ZRDLoaderFull.dll')

print('File exists:', os.path.isfile(Loader))

clr.AddReference(Loader)

(from your PythonNETZOSConnection.py)

Found OpticStudio at: %sc:\program files\zemax opticstudio

Connected to OpticStudio

Serial #: XXXXX (my SN)

(from my lines)

File exists: True

Traceback (most recent call last):

... in <module>

clr.AddReference(Loader)

FileNotFoundException: Unable to find assembly 'C:\Users\...\ZRDLoaderFull.dll'.

bei Python.Runtime.CLRModule.AddReference(String name)

Hi Elisavet,

I'm sorry about the delay. I got this answer from a senior developer here, and am passing it on to you and the Zemax community verbatim:

You cannot easily pass enum values like RayStatus in the DLL (at least the way it was originally setup). In the DLL, the RayStatus, which has the flag for WaveShifted, is converted to an bit-wise integer where all the statuses are consolidated to a single value. To “decompose” the integer into the string value for the RayStatus, you can cast the integer back to an enum. Here is an excerpt from doing this conversion for Matlab, but the process will be identical for doing it in PythonNET:

The Status value comes from the IZRDReaderResults interface and the RayStatus enum. The enum itself is stored as a bitwise signed integer and the actual logic to summarized the RayStatus enum uses a bitwise shift operation to “sum” the integers together to get a “final” status. Since Matlab doesn’t handle enums well, the enum itself is cast to it’s integer value to pass back to Matlab. If the RayStatus is stored in the variable status, then you can use the following code to convert from the integer to the RayStatus:

RayStatusEnum = ZOSAPI.Tools.RayTrace.RayStatus;

System.Enum.ToObject(RayStatusEnum.GetType, status(5))

If status(5) has an integer of 38, then you will get the following output:

Reflected, Transmitted, GhostedFrom

Hi Elisavet,

Thanks for your follow-up here!

After speaking about your comments with someone from our Dev team, I think we can provide some clarifications. First off, you are right -- running the DLL will have all segments of the ray saved serially. However, while Python does have some inefficiencies in looping thorough large structures of data (hence the creation of the DLL in the first place), the usage of Numpy itself will also leverage some other compiled C++ code. So, while there might be some benefit to having a wholly-contained DLL to read through and filter out the data you're interested in (ray segments undergoing downconversion), the difference between that and using the numpy array probably won't be that noticable due to the processing offload done by numpy.

Again, this provided DLL won't be able to do filtering from within the execution -- that would have to be written and compiled on your end for that kind of functionality. However, given how the read data is structured in Python, we think there should be a benefit in the execution of this ray reading procedure with the DLL and numpy library.

Please let us know if you have any further questions or if I've misunderstood you in any way! Thanks again for your time here.

~ Angel

For the above issue, I resolved it the following way: I read the statusData within the loop and got the indices of the downconverted rays with an np.where(statusData==1536). I didn't need to loop over anything. I used np.add.at() to calculate the total absorbed flux after binning the rays in self-made voxel grid. The way I am doing the binning is pretty simple, and perhaps I need to rethink how to do this. 

A very important think to note, is that in order to process all the rays outside the loop you need to increase the number of segments to the actual number of segments, that's roughly 30*no_rays, otherwise you lose info.

I would have posted the code, but don't see a button for code insertion. It's been somewhat frustrating.

Another problem I am having now is that this is using NET, and I don't have access to COM constants. I have written a class using this BatchRayTrace, and I want to write a function that changes the source type. Eg. from Gaussian to Laser Diode, or source Ellipse. How can I do that?

What I've found is this example:

TheNCE = TheSystem.NCE

Object_1 = TheNCE.InsertNewObjectAt(1)

oType_1 = Object_1.GetObjectTypeSettings(constants.ObjectType_SourcePoint)

Object_1.ChangeType(oType_1)

But now I don't have access to constants, because there isn't a COM connection, so that dictionary is not initialised, and I get an attribute error. 

Another point: I can access things from ObjectData in this regime

eg. TheSystem.NCE.GetObjectAt(self.source_no).ObjectData.NumberOfAnalysisRays

Things that are not listed in examples are impossible to find. Can someone explain how the documentation works? I search NCE ObjectData, I get a banch of results. Ok, there are some examples I can extract fields from, but how can I find what I can access through a single tag? I am trying to change the power of the source and cannot find the field through ObjectData.

Edit: Found it, but it got way longer than it should

Another off topic question: Is there a way to return a string or sth to find out what type the object is? Eg. Source Gaussian or source ellipse?

ps. I am using Python