FRI-Energy-Analytics / 2020_summer_fellows

A repository for FRI Energy Analytics fellows
MIT License
5 stars 6 forks source link

Waterfall Chart #71

Open samirabr12 opened 4 years ago

samirabr12 commented 4 years ago

Where is the model used Business/Finance

What is the model A waterfall chart is used in qualitative analysis to show how the original value of something is affected positively or negatively by various factors over a period of time. With this chart, you can see how a certain value is affected cumulatively, and which factors specifically affected it this way. The chart displays the effect on this original value by showing how it changes over time with bar graphs associated with each factor.

How can it be applied to oil We can use this model to see how the cumulative oil production changes during the 12 months, as well as see which features in our dataset affected the cumulative oil production, either negatively or positively.

SsanyaS14 commented 4 years ago

Review 1 by Sanya Srivastava, 06/15/2020, 3:00 pm

Overview This model would make an exceptional comparison tool. We can see that in terms of the aqueous liquid that is compared to each other and cumulative production. Similarly, formation types are compared to each other and increasing production.

The model is the best fit for categorical data. I would recommend looking into the different completion types as well for comparison. However, I do agree with you over how the model is not the best for its estimation, but we should continue to work with it as time goes along with the sprint.

Correlation and error are not particularly stated in this model, so I suggest finding a way to compare the categorical data to the best you can.

Rankings Clear Defintion - A Correlation - D Error - D Potential - B Excitement - B

samirabr12 commented 4 years ago

Revision #1 Comments I agree with Sanya that this model would help with visualization and is worth working with more in order to analyze different variables in the dataset, like completion type. The only concern with this model at the moment is finding a way to compare the variables and how they are correlated with each other since they are not quantitative values.

SsanyaS14 commented 4 years ago

Review 2 By Sanya Srivastava. 06/23/2020, 12:15 pm

Overview I think the model is not going to do so well. The model is a good comparison of the different features but fails to estimate anything as you stated. There is some way to show correlation on the charts between different features but as for the error, I don't believe there is an effective way to show that. Compared to other models that are comparison it does show an interconnectedness to them which is positive, however, it won't have a great application due to the inability to estimate.

Ratings Shown Corr - C Shown Error - D Potential - C Interconnectedness - A Application - D