**This file provides instruction to run the provided source code files**
1.**Source code**
The source code provided here is linked from a github repository. You may download the code directly from this osf project or from the github link: https://github.com/tuanlemau/actor_external
The provided source code is organized as follows:
root folder
+dev: contains main source codes
+data: contains data files
2.**Setup**
The data was analyzed in Python 3.6.10 and Jupyter Notebook v5.5.0.
The most straightforward method to execute the provided code is to install the latest version of Anaconda (in either Windows or Mac) from anaconda.com
Use Pandas version 1.0.1, as future versions have been reported to break the code.
Once downloaded, simply launch Jupyter and navigate into the root folder. This is where this instruction should be found.
Software library versions
* pandas v1.0.1
* numpy v1.19.2
* seaborn v0.11.0
* scipy v1.5.2
* sklearn v0.23.2
3.**Running the code**
Navigate to /dev folder to access the source codes
There are the following main source code files
1. actoracting_reliability_specificity_analysis-Analysis.ipynb
2. actoracting_reliability_specificity_analysis-Multiverse.ipynb
3. inductive_clustering_analysis.ipynb
These cover the main analyses in the paper
The remaining files in /dev are supporting codes to be imported by the two main source code files
There are a number of options in the main source code files to experiment and generate analysis. The readers are welcomed to experiment with the options.
The code should run out of the box. If there are any issues, please feel free to contact the author at tuanlemau@mit.edu
***data folder***
To avoid breaking the source code, do not modify any files in data folder.
The file survey_data.csv located in data/core_data contains the participant rating data.
The file survey_questions.csv located in data/core_data contains the action units coded for a given portrayal. These can be linked to the source books using the id variable which corresponds to index in the Stimuli database.xlsx hosted on this osf project.
**Other notes**
1. No non-standard hardware required
2. No other required dependency apart from Anaconda package
3. Demo: all required data is provided and will automatically by loaded when executing two main source code files
4. No other instruction for use is required or provided. If issues are encountered, please feel free to contact the author at tuanlemau at mit dot edu OR t_lemau at mit dot edu
