Releases: nabeel-oz/qlik-py-tools
Releases · nabeel-oz/qlik-py-tools
v.6.0
Feature Release
Exciting new capabilities for Deep Learning with Keras and Tensorflow. You can now train and use neural networks for sequence predictions and complex timeseries forecasting.
This release also includes the ability to use Additional Regressors with Prophet, allowing you to model more complex timeseries.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning : Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Deep Learning : Implemented using Keras and TensorFlow. This SSE implements the full flow of setting up a neural network, training and evaluating it, and using it to make predictions. Deep Learning models can be used for sequence predictions and complex timeseries forecasting.
- Named Entity Recognition : Implemented using spaCy, an excellent Natural Language Processing library that comes with pre-trained neural networks. This SSE allows you to use spaCy's models for Named Entity Recognition or retrain them with your data for even better results.
- Association rules : Implemented using Efficient-Apriori. Association Rules Analysis is a data mining technique to uncover how items are associated to each other. This technique is best known for Market Basket Analysis, but can be used more generally for finding interesting associations between sets of items that occur together, for example, in a transaction, a paragraph, or a diagnosis.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts. Now with the ability to use multiple regressors as input.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
For more information refer to the usage section.
Change Log v.6.0:
- Sequence predictions with Keras as documented in Keras.md.
- Additional Regressors for Prophet as documented in Prophet.md.
- New docker image published.
- Sample apps for the new capabilities.
- Improved simple forecasting app for Prophet.
- Better logging. You can see more details for each function call in the SSE logs.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.5.1
Maintenance Release
Exciting new capabilities for Named Entity Recognition and Association Rules Analysis.
Changes for issues #29 and #34.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning : Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Named Entity Recognition : Implemented using spaCy, an excellent Natural Language Processing library that comes with pre-trained Neural Networks. This SSE allows you to use spaCy's models for NER or retrain them with your data for even better results.
- Association rules : Implemented using Efficient-Apriori. Association Rules Analysis is a data mining technique to uncover how items are associated to each other. This technique is best known for Market Basket Analysis, but can be used more generally for finding interesting associations between sets of items that occur together, for example, in a transaction, a paragraph, or a diagnosis.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
For more information refer to the usage section.
Change Log v.5.1:
- Additional parameters for Prophet as listed in updated Prophet.md
- Fix for machine learning when none of the features require scaling
- New docker image published
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.5.0
Feature Release
Exciting new capabilities for Named Entity Recognition and Association Rules Analysis.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning : Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Named Entity Recognition : Implemented using spaCy, an excellent Natural Language Processing library that comes with pre-trained Neural Networks. This SSE allows you to use spaCy's models for NER or retrain them with your data for even better results.
- Association rules : Implemented using Efficient-Apriori. Association Rules Analysis is a data mining technique to uncover how items are associated to each other. This technique is best known for Market Basket Analysis, but can be used more generally for finding interesting associations between sets of items that occur together, for example, in a transaction, a paragraph, or a diagnosis.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
For more information refer to the usage section.
Change Log v.5.0:
- Named Entity Recognition capabilities using spaCy
- Association Rules with Efficient-Apriori
- Sample apps and documentation
- New docker image
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.4.4
Maintenance Release
Additional fix for regression models.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.4.4:
- Fixed the numeric return type failing for the prediction function.
- New docker image released with the changes.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.4.3
Maintenance Release
Fixed an error in the return type when getting predictions from a regression model.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.4.3:
- Fixed the prediction return type when using regression.
- New docker image released with the changes.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.4.2
Maintenance Release
Updates to initialisation to correct for new packages.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.4.2:
- Fixed versions for hdbscan, fbprophet and scikit-learn in initialisation.
- Qonnections 2019 workshop materials added to the project.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.4.1
Maintenance Release
Fix for hdbscan incompatibility with new version of scikit-learn.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.4.1:
- Updated setup to use scikit-learn version 0.20.3. This avoids incompatibility issues with hdbscan.
- Made changes for pending deprecation of joblib from sklearn.externals.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.4.0
Feature Release
The Prophet function now supports calls from the Qlik load script.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.4.0:
- Updated the Prophet function to allow usage from the Qlik load script.
- Updated sample app to provide example of Prophet being used from the load script.
- New image published to Docker Hub.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.3.9
Maintenance Release
Updated initialization script.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.3.9:
- Updated initialization to use the fbprophet 0.4.post2. This fixes an incompatibility with a new version of numpy.
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.
v.3.8
Maintenance Release
Updated initialization script.
This release includes:
- Supervised Machine Learning : Implemented using scikit-learn, the go-to machine learning library for Python. This SSE implements the full machine learning flow from data preparation, model training and evaluation, to making predictions in Qlik. In addition, models can be interpreted using Skater.
- Unsupervised Machine Learning: Also implemented using scikit-learn. This provides capabilities for dimensionality reduction and clustering.
- Clustering : Implemented using HDBSCAN, a high performance algorithm that is great for exploratory data analysis.
- Time series forecasting : Implemented using Facebook Prophet, a modern library for easily generating good quality forecasts.
- Seasonality and holiday analysis : Also using Facebook Prophet.
- Linear correlations : Implemented using Pandas.
Change Log v.3.8:
- Fix for initialization error introduced by fbprophet 0.4
This zip archive only contains the files needed to deploy the SSE. To get the sample apps download the full source code above or get them from the docs.