pladif 1.3

Description:

pladif 1.3

PLADIF: Plots Attrakdiff graphs from CSV files
Attrakdiff is a method to evaluate UX aspects like attractivity, usability, desirability, etc.
Usabilla is a software that is able to get feedback from online customers.
PLADIF is a simple tool that plots attrakdiff plots from CSV file (like those prroduced by Usabilla). It is based on Python, matplotlib, pandas and streamlit. It's a web-app that can be installed locally or hosted in a web server.
A live demo can be found here
The web-app takes Usabilla's CSV files as input, and produces attrakdiff graphes as output.

It produces the three diagrams of the Attrakdiff method:
Diagram of average values

Description of word-pairs

The portfolio presentation

Installation
PLADIF is a web-app done using Python, matplotlib (for the plots), pandas (for the data manipulation) and streamlit (for the web-app). These libraries are way overkill for a such simple tool, but it made my development much easier 😀 !
On Mac or Linux
To install it, you need to have a machine with Python3 installed. You then just need to install the pladif library, with
pip3 install pladif

and that's it ! (ok, it will install a lot of things, specially if you don't use python for anything else).
The right way to do it, is of course to do it in a virtuel environment.
On a fresh Mac, the system will probably ask to install some developper tools first (do it).
On Windows machine
You probably need to install it using Conda, and then install the pladif package.
run PLADIF
To run PLADIF, just launch the runPladif script
runPladif

(if runPladif doesn't work, it means the package pladif is installed, but not added in your path)
On a first run, streamlit will ask for an email, juste press Return (never give your email address to strangers 😉). You then have the following message


⚠️ Press Ctrl + C to stop PLADIF ⚠️



You can now view your Streamlit app in your browser.

Local URL: http://localhost:8501
Network URL: http://192.168.1.18:8501

For better performance, install the Watchdog module:

$ xcode-select --install
$ pip install watchdog

and it means that it works ! It should also open a tab on your web browser, with PLADIF open.
Don't forget to close PLADIF (the server) with Ctrl+C when you don't use it (close the browser tab is not enough)
Use it
It's quite simple. Just drag'n drop your CSV files (from Usabilla) on the left panel, and that's it.
You can change the lang (English or French for the moment, Deutsch should arrive soon), or adjust the interval confidence level.
You can download each image (with the download button below each image; you can choose the file format in the plot options).
TODO

add Deutsch support
integrate all the feedback you may send (just open an issue on GitHub)
add a CSV (or excel) report, with all the data
add a pdf report
add a "quit PLADIF" button ?

Versions

v1.3: add resolution menu (in figure option) to choose the image file size
v1.2: add support for Excel files (the 1st row contains the header of the column)
v1.1: add summary of the files
v1.0: PLADIF is now mature enough to have a 1.0 version !
v0.5: correct bug in pair-word figure
v0.4: add various image formats for the download (jpeg, tiff, pdf, svg or png)
v0.3: display confidence intervals in the tables
v0.2: plot confidence intervals (based on Student's t-distribution, that is probably different that the one used by Attrakdiff, but I don't know there is no documentation about it there)

I hope it will be useful
If PLADIF is useful, buy me a beer 🍺 !
Disclaimer: I am not affiliate to Usabilla nor Attrakdiff. This is a simple python tool for that. It uses matplotlib for the graphs and pandas for manipulatin the data (I am not a pandas expert, and probably some code that be done more efficiently with the adequate pandas methods). Streamlit is used for the web app. It is maybe not the best choice for PLADIF, but I wanted to try it!