After several years of communication with my friend A. it was revealed that constructiveness of conversation can be achieved only if he is in a good mood. On the reverse side if he is in a bad mood discussion not only won’t bear fruit but it can have even negative consequences in terms of patience and reasonableness.

The lack of understanding of the current process created difficulties in communication and required solutions.

The purposes of this application

• Try to explain my friend A’s past actions based on his moods track (it is also possible to see the track of predictive variables if necessary);
• Predict his moods in current/or nearest future and adjust possible patterns of behavior on meetings.

After investigation it was revealed that following variables influence on A’s mood:

•   Wind Speed (m/s) at the opposite point of the globe;
•   Days Untill annual Kayaking;
•   Moon Phases;
•   Is he in a pizzeria Underdog?

Mood represents on scale from -10 to 10 (where -10 – extremely aggressive behavior, 10 – polite and very sympathetic). Since Negative Influence on his mood in pizzeria Underdog is so significant that neutralizes all the other variables => it was withdrawn from analyzing historical trend. And if you try to predict future A’s mood while he is in pizzeria Underdog – application automatically puts -100.

Model used for the analysis - Regression model

Link on Github: GitHub

Application Link: Shiny Application