“We are on the final slope!” – Mike Ames.

As the semester progressed, Mike and his one offs became more and more hilarious. Yes, the final slope has arrived. The information and skills I have learned are amazing. I’m sure this is still just the tip of the iceberg when it comes to R and SAS.

I feel accomplished and I am looking forward to working with these tools in my future classes and career.

What stands out the most to me about R, is that there always seems to be a better package… ever making it more convenient to use. The syntax isn’t easy at all… but the packages compile more steps into itself.

Chapter one of graduate school is coming to an end… chapter two awaits.

I question whether I like R or SAS more. We have spent equal time learning each. I like how clean and neat SAS is, however, R is more powerful, yet a little chaotic. I believe SAS and R have their respective roles in the professional world.

I think if you want neat presentations and are not concerned with cleansing mass amounts of data SAS is the way. If you are looking for the opposite then R is your choice. The packages in R make tasks much easier than they would be in SAS. Yet, I am not completely sold on R.

This is just my humble little opinion as I have only been coding for 15 weeks. Maybe my perspective will change as time passes and I progress in my Masters and future career.

Janitor is a pretty neat tool, as is e1071!

Text analysis is super fun. Again, I am blown away by the power of R. The more I boil down the libraries and understand how they interact the larger the picture becomes. Text analysis must be a critical part of customer surveys when they are not measured on numeric value.

The ability for Text analysis and word clouds as well as word matrix’s to pull words and find connection to other words and assign sentiments in insane. Again, these tools make it easy to understand large quantities of text data and boil it down.

If I worked for a larger contractor, it would be intersting to run a text analysis on all the emails from clients. It would be interesting to see what words attach to what sentiments and how they relate to each other. For example, budget, upset, beautiful… etc.

As we learned MACRO’s in SAS, R has Functions. Functions are a time saving powerful tool. It’s ability to run multiple iterations of the same complex code is great.

I still can’t decide which software I like more. I like how clean and pragmatic SAS is, however, R allows you to do a lot more, more simply… not necessarily clean looking but the open source nature allows R to have some powerful abilities.

As we move forward to linear regression, clustering and correlations. I see now that I could build a model for estimating build costs. Designing a predictive model for new construction would be amazing. It would stream line the estimating process by allowing an estimator to put in specifics regarding, sq ft, beds, baths, porches, quality of finishes… etc and create an estimated cost. As of now, most estimates are based off of generic square foot numbers, or actual hard math.

Not really, just a pun on regression. So, regression modeling is much easier in R (compared to SAS). So easy it feels like cheating. However, I do like the statistical out put of SAS more. Maybe I just don’t know how to output the data in R the best.

I am excited to use the new tools I’ve learned on some real world data. A friend of mine is going to let me look at some data he gathered for his company. I hope to be able to use the skills I have learned the past couple weeks in R, and tap into my memory bank of SAS to look at his data. It will be fun to actually look at the data in both SAS and R.

resample_partition is a powerful tool. You can easily partition your data on a 70 30 basis to train and test to run your regression analysis. I look forward to working with this more.

R continues!

It is starting to make more sense. The pipe function is stellar. It’s a great way of passing on all the code you have created into your next step. It is extremely useful when filtering, grouping, aggregating data for ggplot.

I think knowing pipes, and having the ability to make comparison plots with ggplot would have been useful when comparing year over year financial returns. Or rather simply put, P and L over years.

It’s hard to always know how to correctly filter, and clean the data. The more I do it the easier it becomes but every week when something new is introduced it becomes a little harder. But hey, I guess that is the name of the game when it comes to coding!

R is pretty awesome. It is nice having the foundation in coding. What I learned in SAS does not directly translate into R. However, having an idea of how to structure code make it easier to understand, and understanding the basic principles behind functions and processes.

I really like how you can work with matrixs and vectors. It’s neat to see the different ways you can manipulate the data. The packages and libraries seem endless. It is going to be interesting to learn them on and determine which are the best for the application I need.

Once you know the process you want to run, R seems more user friendly.

This week we transition to R. At first I thought this was going to be a quick and easy transfer… so far it hasn’t. To keep all things equal, I haven’t spent much time with it since I was finishing up the SAS midterm.

At the conclusion of the SAS midterm I have a better understanding of the SAS process. The program is extremely capable and I hope to transfer the SAS skills to R. I wonder if R is as dependent on syntax and process? The hardest part of SAS is remembering when you use a when, by, or other statement.

Thats it… after this post we will be onto R. I took a look at R for everyone and although, it is completely new to me, after working with SAS I recognize the processes of it. That is, you import, structure, cleanse, sort, and manipulate!

The week leading up to the midterm we learned how to create regression modeling. This is an extremely powerful tool. It really is beginning to take shape. I am not a master of SAS yet, however, I wish we were going to continue with the program for a few more weeks. I think with more experimentation and working with the program I will really get the hang of it. It is fun to see how SAS operates and how it is correlating to our statistics and probability modeling class. Sometimes I don’t fully understand what we are doing in SAS until I get some more information in statistics.

I hope my experience with SAS is going to make coding in R easier.

If I had enough data from work or other contractors, I wonder if I would be able to create a regression model for estimating build cost. It would have to account for sq ft, cost of materials, level of quality, finish, etc… It would take quite a bit of data mining to get that started.

REGRESSION! Is the talk. This week I learned Proc Reg… rather proc reg owned me! I need to spend more time with this proc, however, I see how powerful it is.

Understanding Proc Reg will help me to forecast financial costs and material costs. If I have enough data I should be able to upload order files and use cost, sq ft, material to identify how much I will need for projects based upon sqft. If I was aware of this before, I would have designed better tracking methods.

…I think I sense a side project.

I will need to go through Proc Reg a few more time and experiment with some alternate data to get a better understanding of it. I was having trouble understanding when we assigned the predictive nature to quality and wasn’t able to adjust my RSME at all. I must be missing something and need to take some time to go back and look.