14
Aug

Getting Started with Orange 08: Add-ons


By now, you know Orange quite well. But how about if you want to do some specialized research in, say, bioinformatics or text mining? Add-ons are coming to the rescue. You can find add-ons under Options – Add-ons. This will display a package list from where you can install the add-ons you want. Say we want to go with bioinformatics. Select the package you wish to install and press OK. Now restart Orange and you will see an additional widget set on the left. Ok, let’s explore what’s in the add-on. There are some visualization widgets, some specialized data analysis techniques, and a few widgets that offer access to external databases. One such widget is GEO Data Sets that offers access to a large repository of gene expression profiles. There are a few thousands of data sets you can load from the widget. Say, I’m interested in researches on smoking. We first filter the data on this keyword. Now let’s select the small data set. This one looks nice. First, we want to see what kind of data we got. These are gene expression data for five smokers and five non-smokers. We have gene expressions in columns and samples in rows. Now we want to visualize these data to quickly check if there’s any difference between smokers and non-smokers based on their gene expression profiles. There are so many features to choose from so how do we make sense of it? Multi-dimensional scaling will help. This technique projects multi-dimensional data to a 2D space. We see there are two nice groups, blue for non-smokers and red for smokers. There is one sample from a smoker which is outlier. This could be very helpful. Even if we’re not molecular biologists we can see that smoking leaves a trace on a cellular level. But enough biology for now. Multi-dimensional scaling and other embedding techniques are a great way to understand the underlying patterns in any kind of data. We will talk more about 2D embedding in the upcoming tutorial. As for today, we have learned how to install add-ons in Orange and how to use new widgets in combination with the existing ones.

Tags: , , , , , , , ,

16 Comments

  • Sriharsha B S Prasad says:

    Nice explanation!!!!but I am unable to install add ons through options->Add ons…I am getting error..its not showing any widgets like you displayed there

  • mauhamad annouss says:

    hey guyss .. love ur great software .. anw we are waiting for more videos as soon as possible please 😊

  • Amit Solanki says:

    Waiting for more videos.. Great explanations, would be great if you guys went in more depth for each classification technique

  • Shunsuke says:

    Thanks for these valuable video clips!

    Orange is much much better than Weka.
    I really fell in love with it.

    Where can I find more tutorials and exercises to learn machine learning and data analysis by using Orange?

  • Tarcísio Bruno Oliveira says:

    Waiting for more tutorials. This is an awesome playlist! Thanks.

  • Adi Bandaru says:

    very impressive. liked the idea of python as background engine for Orange. what is the feasibility to use in enterprise level. any metrics will be appreciated.

  • Rafael Calastro says:

    This ORANGE is quite amazing! Lightweight, powerful and easy to learn!
    I´ve subscribed the youtube channel and I´m wondering what is coming next…

    By the way, Ajda is astonishing beautiful and very intelligent… what a combination!
    Thanks…

  • lavorboj says:

    Hope to see more videos and updates on http://orange.biolab.si/toolbox/ soon. 🙂
    Orange is amazing

  • Olivia Belisle says:

    Hey! Does anyone know how the Infrared Add-on works??? I cannot find information on it anywhere… We hit a road block in our data analysis!

  • Karla Sarai says:

    Hi!
    When I try to install network add-on, the program fails and Send a message: "An error occurred while running a subprocess.
    Comand failed: python-m pip install orange3-Network exited with non zero status" What can I do? Thanks!

  • Joshua Wojnas says:

    Hi I like the neuronetwork widget and that it uses scikit learn now I seen scikit learn auto ML I have seen some great things are done with it and google deepmind alpha zero. It also had a stack based neuro network that might be able to be made with the right formated data and orange widgets there are some great python tutorials out there on it.

  • O Ruk says:

    Hi
    I installed Orange for windows last week and when I ask for the add-ons list the following warning pops up:
    Could not retrieve package list. <urlopen error [WinError10060] … >
    Any idea?
    Thanks

  • Said Benaissa says:

    Clear explanation. Could you please explain how to use time series add-ons with some examples? thanks

  • Bhrigu Chadha says:

    Need help! 1.) Package list cannot be retrieved 2.) I have installed two addons but the widgets are not appearing.

  • aurora cooper says:

    it won't let me install add ons. it says: Insufficient permissions to install add-ons. Try starting Orange as a system administrator or install Orange in user folders.
    can someone please help me?

  • Berns Buenaobra says:

    Good for practice and student work but can it crunch really big data – let's see will give it a try.

Leave a Reply

Your email address will not be published. Required fields are marked *