Hot answers tagged

6

As you mentioned, Orange is a data mining software developed by the University of Ljubljana. It can be used for developing and testing machine learning models as well as conducting exploratory data analysis and visualization. One of the unique features that makes Orange "special" is its simplicity and ease of use. This is because components in Orange are ...


5

This was an error in Orange 3.3.11, and fixed in release 3.3.12. Update to the newest release to get rid of the error.


3

The error message shows exactly what you need to do: Orange requires Python>=3.4 You have to specify a Python>=3.4 version (with orange3 itself installed) while installing orange3-prototypes. I'm not a Mac user, however, in Windows, Orange3 installer will automatically install a Python 3.4 if there's no compatible Python available.


3

There are three ways to install Orange add-ons: The Options -> Add-ons menu of Orange. Use pip: pip3 install orange3-prototypes Install from the source: Terminal commands: git clone https://github.com/biolab/orange3-prototypes.git cd orange3-prototypes python3 setup.py install


2

The easiest way to install Orange's addons is through the application itself. Open Orange, in the menu click Options -> Addons. In the popup window mark Orange3-Prototypes and click OK. Note, that doing so you will get the latest version that is published on PyPi. If you would want to install the bleeding edge version directly from github — assuming that ...


2

So you have two independent variables. iOS version and release name, which is actually a type. Both are categorical and only one has any predictive power, because iOS version you are trying to predict does not appear in your training set. Thus, effectively, you are predicting based on the release type only and only the "release" category is actually ...


2

Use Feature Constructor like so: and X1 will be the name of your new feature (you can obviously change it)


2

This is a known issue: https://github.com/biolab/orange3/issues/3193 It is caused by a bug in Qt 5.9.6 (QTBUG-16252). You should downgrade Qt to 5.9.5. To do that open the "Orange Command Prompt" from the Window 8.1 'Start Screen' (start menu) and type in conda install qt=5.9.5


2

Yes, this is what the Save Model widget does under the Model tab. Create your model, then click save model and your model will be saved to a pickle file. Then just load your pickle file in Python. Here is a like to a similar question on Stack Overflow: https://stackoverflow.com/questions/45598627/orange-save-model-in-python-script


2

Meta variables are meta data, data about data, not used for statistical inference. Features or variables or attributes are the measured inputs of the problem domain, the independent variables. The target variable is the dependent variable or the measure we're trying to model or forecast. Not all problems can be or need to be formulated in such a way. Orange ...


2

Yes, Orange does not handle very large data sets very well. In fact, if you try to load in something like the file size that you are talking about (gigabytes of data) it will crash. When ever you want to create a large scale neural network, you will need to use packages like Tensorflow and Keras. You can not do this in Orange.


2

All release notes for new versions of the Orange Data Mining software can be found by looking at the release page on its GitHub page here. Then just navigate to the most recent release and click on the latest branch. This will display a window similar to the following where you can view all of the changes that were made (both enhancements and bug fixes) in ...


2

Weighted linkage probably does not mean you get to specify weights of features (build the distance matrix yourself!) Instead this most likely refers to the well-known weighted group average strategy you will find in most textbooks often called WPGMA. There are two different definitions of "average", so this is likely simply the "other" average linkage.


1

Maybe you could use Orange3-Text add-on, widget Preprocess Text, Tokenization > Regexp. The source code indicates it's a Python regex, so you might be able to use a regular expression pattern such as: (?ix) # ignore case, ignore comments and whitespace in this RE (?<=age:\s) # preceded by 'age: ' .+ # characters you wish to match (?=...


1

Yes, orange has regex abilities as seen here. If you're unfamiliar with regex here's a link regarding regex with numbers. The idea is pretty straight forward. Use regular expression to find all numbers, then replace the numbers with something like a blanks string a placeholder word like "NUMBER".


1

ARI stands for adjusted Rand index and AMI stands for adjusted mutual information. They are metrics for clustering. Remark: ARI might take negative values. The AMI takes a value of 1 when the two partitions are identical and 0 when the MI between two partitions equals the value expected due to chance alone.


1

Firstly, I am assuming that you are using Anaconda Navigator, and are trying to use Orange Canvas (the GUI method) If the package has been installed correctly (from within Anaconda, using the Environments tab) you will see the following when opening Anaconda Navigator: So, there is not a Windows desktop or start-menu icon; instead launch Orange from ...


1

I just realised I'd missed the 'colour' widget. By connecting the data table to the 'colour' widget, I could set what colour I wanted different data sets to be. Then I could attach the colour widget to my scatter plot, and BAM, the colours I wanted. I needed to make sure the scatter plot widget was connected to both the colour widget and the data table. ...


1

Your workflow is correct and you are doing a type of transfer learning where only the last layer is fine-tuned. Image embedding widget calculates features that are the activations of a penultimate layer of Inception-v3, VGG or Painters network. All of those networks are trained on the different set of images. VGG and Inception-v3 are trained on ImageNet ...


1

Python Script widget looks like it might fit your needs, where, according to the docs, a code like this should give you the number of rows in a data set: n_rows = len(in_data)


1

Yes, Orange has a Dimensionality Reduction Widget under the Unsupervised Learning tab. Click PCA (Principal Component Analysis)


1

Finally I found the answer. In elastic net the cost function is written as $J(\theta)=MSE + r\alpha\sum_{i=1}^n \vert\theta_i \vert + \frac {1-r}{2}\alpha\sum_{1=1}^n \theta_i ^2 $. Here $r$ is the ratio, $\alpha$ is the hyperparameter and $n$ is the number of features. The ratio slider controls $r$ and $\alpha$ slider controls the value of hyperparameter. ...


1

Can't you do something like this? (or am I missing something?)


1

As r.kfr said: Import Images works with folders, not individual images. Place all images in a folder and select it for import. Note that if you have images in different folders, Orange will consider each folder name as a class label for images. Alternatively, you can create an Excel/csv file and define the path to the image. Mark it with meta type=image in ...


Only top voted, non community-wiki answers of a minimum length are eligible