Conversational Data Browser

Level 4 Individual Project

main.py is the file that should be run to start the application. The conversationalbrowser/ folder contains the user interface files under the folder ui/. The folder also contains three python files. data_manipulation.py is how the application extracts required data from the dataset, graph.py which controls the displaying of the graphs, and the model which contains the classes that hold the various data.

Build instructions

Requirements

• Python 3.7 and above. May work for Python 3.5 and 3.6 but has not been tested.
• Packages: listed in requirements.txt
• pip3 installed. This comes automatically using Anaconda on Windows.
• Default file explorer. Windows, OSX and Ubuntu already have this installed by default.
• Tested using an Anaconda virtual environment on Windows 10 (may need to install Anaconda) for ease of use.
• Tested on Ubuntu Linux 20.10. Should work on older versions.

Linux

• Update pip using either pip3 install --upgrade pip or pip install --upgrade pip on what is being used.
• Ensure you have pip3 installed and Python 3.X. This has been tested with 3.7 and 3.8.
• Download the files using either the code download as zip or git clone.
• In terminal, same folder as the main.py and requirements.txt, run pip3 install -r requirements.txt or pip install -r requirements.txt (depending on version).
• Then do python main.py
• To edit the user interface using QtDesigner, do cd /usr/lib/x86_64-linux-gnu/qt5/bin/ to get to the folder containing QtDesigner and then run ./designer.

Windows

• Update pip using pip install --upgrade pip
• A couple of ways. Recommended way is to have Anaconda installed.
• In the Anaconda Prompt navigate to the folder containing requirements.txt and run pip install -r requirements.txt. Filelock installation error may occur but shouldn't affect the running of the program.
• Then do python main.py
• To run the QtDesigner to change the UI, designer.exe will be under ....Lib\site-packages\pyqt5_tools

Test Steps

• Testing is automated if the code is hosted on Github, using Github Actions.
• Testing can be run locally by executing pytest --cov=./ tests/ in the root directory.

Documentation

Docs for the code can be found in the repository as a html file under html/index.html

Troubleshooting

• If there is a marshal data too short error, try pip3 install --upgrade --force-reinstall <package that is not working on>. If this fails, it probably means your pip installation is broken. Follow the relevant uninstall/reinstall guides for your system.
• Mac systems might have some issues with pip installation of packages. The workaround is to install the packages manually and ignore the version number.