Troubles with additional dependencies

Hi Everybody!

I am facing problems with deploying a python serverless function to azure. I am not able to find solution in the Internet.

My biggest problem is I don't know how to install additional dependencies (I have libraries created by myself I don't want to publish in Python Package Index). I tried everything I found in the Internet (except the Kudu because I am not able to run Kudu for linux python functions). Everybody in the internet suggests to make a wheelhouse directory and put *.whl files in it and then put --find-links as the first line of requirments.txt file. But this solution doesn't work for me. During publishing the function I get an error about that my wheelhouse directory doesn't exists or no dependencies are installed.

My other problem is that I have in my requirments.txt file:
```
azure-functions==1.0.0b4
azure-functions-worker==1.0.0b5
cycler==0.10.0
decorator==4.4.0
grpcio==1.19.0
grpcio-tools==1.19.0
kiwisolver==1.1.0
llvmlite==0.28.0
matplotlib==2.2.2
networkx==2.3
numba==0.43.1
numpy==1.14.2
opencv-python==3.4.0.12
Pillow==6.0.0
protobuf==3.7.1
pyparsing==2.4.0
python-dateutil==2.8.0
pytz==2019.1
PyWavelets==1.0.3
scikit-image==0.13.1
scikit-learn==0.19.1
scipy==1.0.1
six==1.12.0
thinning-py3==1.2.3
```
but I am not able to install it because `thinning-py3` requires `numpy` but it cannot be found.

Any advice is appreciate.

Best regards
Paweł

Start by reproducing the Azure Functions deployment using the requirements.txt shown in the report, including the wheelhouse setup and the thinning-py3 dependency. Determine whether the failure is in dependency discovery, private-package installation, or NumPy resolution; done means the function deploys with all listed dependencies installed.

Written by the indexing model from the issue text.