oranchada

Orange workflow add-on for processing Raman spectra. Based on ramanchada2.

• Poster: OPEN-SOURCE FOR RAMAN SPECTROSCOPY HARMONISATION
• ORANCHADA visual guide (Widget usage examples)

Installation

UPDATE: This guide used to suggest using the standalone installer of Orange. It is now recommended to use the portable version of Orange instead. This avoids some incompatibilities. Note as well that Oranchada requires Orange 3.37 or newer.

1. Depending on your operating system:

   1. Windows: Download the Portable Orange ZIP from https://orangedatamining.com/download/#windows and extract it to some place from which it will be convenient for you to start it. If you are not sure about the place, just extract it to your desktop (it will be possible to move it later if you change your mind). To run Portable Orange, open the shortcut. On the first run, there might be a security warning. In this case, unselect Always ask before opening this file and click Open.

   

   

   

   2. Linux: Use the instructions in https://orangedatamining.com/download/#linux. However, we strongly recommend creating a separate environment (i.e. not the base one) if using Conda or a venv if using pip.

   3. macOS: Not tested; the instructions from https://orangedatamining.com/download/#macos should work. Feedback welcome.

2. Once it's installed and running, open Options ➡️ Add-ons. Note the "Orange Update Available" notification in the lower right corner that may show up at times. Do not click download. Refer to the updating section below for more information.

   

3. Click the Add more... button.

   

4. Type oranchada in the Name field, then click Add.

   

5. Scroll down the list until you find oranchada, then click the box next to it so that a check mark appears and Install is displayed in the Action column. Then click OK.

   

6. Orange will then proceed with the installation. When ready, it will suggest to restart itself. Click OK.

   

7. Once Orange restarts, you should see an Oranchada Easy and Pro categories in the left panel. Click them to reveal all widgets.

   

8. To verify the Oranchada installation:

   1. Click the Load Test Spectra widget or drag it to the workflow area to the right.

      

   2. Right-click on the widget in the workflow area and select Open or simply double click the widget.

      

   3. Scroll down the options on the left until you reach the filenames section and select an arbitrary filename from the list (e.g., the first one).

      

   4. Scroll the options back to the top and click the Plot button. If Oranchada was installed correctly, you should see a spectrum visualization on the right.

      

Manual installation

If your network connection has very tight security that severely restricts what sites you can access, the usual installation method via Options ➡️ Add-ons might not work for you. It may be possible to circumvent the problem by installing oranchada manually:

1. Open https://pypi.org/project/oranchada/#files and download the "Built Distribution" .whl file.
2. Open Options ➡️ Add-ons then drag & drop the downloaded .whl file to the add-ons window.
3. Proceed with the general instructions above, starting from step 5 ("scroll down the list until you find oranchada").

Updating

1. Open Options ➡️ Add-ons. If there's a new version of oranchada or Orange itself, they will be listed at the top, with your installed version and the latest available one indicated (e.g. you have 0.0.6 installed and 0.0.7 is the latest available version that you can update to).

   

2. Click the box(es) next to the available updates so that a check mark appears and Update is displayed in the Action column, then click OK.

   

3. If Orange itself has an update in addition to oranchada, take note that:

   1. An additional warning will be displayed. Click OK.

   

   2. IMPORTANT: In case the upgrade fails, click OK and repeat the steps above—it should work on the second attempt. Do NOT close Orange here without repeating the steps! If you do this, you will most likely not be able to start Orange again and will have to install it anew from the beginning, possibly losing some of your settings.

   

User guide

ORANCHADA visual guide (Widget usage examples)

Troubleshooting

Windows cannot find 'Orange\pythonw.exe'

You may encounter the following error message when trying to start the portable Orange:

A possible workaround is to open the Orange directory and then the Scripts one and start orange-canvas.exe from it. You may also need to restart Orange manually after installing or updating Oranchada or other add-ons.

api-ms-win-core-path-l1-1-0.dll missing

You may encouter this or a similar error on Windows 7 and earlier. There are some workarounds, but since they require messing with important Windows components, we cannot recommend them to the casual user.

For developers

Always use a virtual environment for better reproducibility. Poetry may be introduced in the future, but currently it is recommended to use Conda.

Start by cloning the repo:

git clone https://github.com/h2020charisma/oranchada.git

or, if you have write access too:

git clone [email protected]:h2020charisma/oranchada.git

Conda

Miniforge is recommended.

Quick environment setup:

conda env create

This will create an environment named oranchada with Orange and oranchada installed in it. Oranchada will be installed in editable mode (a.k.a. pip install -e); restart Orange to have your local changes to the code picked up. To run Orange in the environment (don't forget to activate it first with conda activate oranchada) use orange-canvas.

If you need to also change the code of some of the dependencies, e.g. ramanchada2, install them in editable mode as well (this should override the previously installed dependency from package). Again, don't forget to do this from within the activated environment.

pip install -e ../ramanchada2

The quick setup above uses the provided environment.yml file. You can create a basic environment yourself too:

conda create -n oranchada -c conda-forge -c default pyqt orange3
conda activate oranchada
pip install -e .

To develop or test with the latest ramanchada2 from Git, an alternative conda environment is also provided for convenience:

conda env create -f environment-dev.yml

Python venv

Make sure to use an appropriate Python version, typically what Portable Orange is bundled with (you can also check environment.yml in this repo). For POSIX systems, pyenv is convenient. For Windows, there's pyenv-win.

cd oranchada
python -m venv .venv/oranchada
source .venv/oranchada/bin/activate
pip install -e .

For other POSIX shells and Windows, see https://docs.python.org/3/library/venv.html#how-venvs-work

Acknowledgements

🇪🇺 This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 952921.