What I've Learned So Far

May 08

Well I'm bouncing around distracted, because I'm waiting to start DAND Term 2! It seemed high time to do another 'what I've learned' post - I missed the last month because I was on vacation.

So here's what I've learnt:

• Numpy and Pandas:
• So much easier to manage databases with pandas than with straight Python!
• That even though I still have MUCH to learn, I can functionally use numpy and pandas to do some neat things
• Googling seriously is your friend! So are ppl who write nice documentation and good stack overflow explanations
• I really do love this data analysis thing and can spend WAY more time on projects than is required


• Bootstrapping:
• Learning this was a truly 'mind blown' experience for me - while in some ways it is all about traditional statistics, but some much of "unlearn what you have learned" - in a good way
• The trend in statistics is to start showing how all my go-to statistical measures might not be as effective as we thought - p-values, r², null hypothesis testing! What are you all doing to me?! (I still need to write a post on all of this)
• How to estimate the parameters of a popluation!! The fact that this previously unknown entity is just amazing to me
• How to calculate a p-value based on my hypotheses - important stuff here.
• I have once again become 'that student' - the person that has to know all the intricate details of why and pesters her instructors with questions pre-written down, waiting for them to be available!


• Mulitple linear regression:
• The stats brain is alive and well!!
• The stats that I love relate to machine learning! I'm so looking forward to being able to get more into machine learning for modeling data
• How to use multiple regression and pandas to complete ANOVAs (I do like ANOVAs)
• That with the computational advances (and less focus on all of the technical math behind it), in the three months of Term 1 we are about half way through my 3rd year statistics course


• How to write READMEs:
• The intimidating is now no longer so!
• I now need to go and write proper READMEs for all of my projects on GitHub


• Kaggle exists:
• People who want to host datasets so that they can sit there for me to find and explore make me happy
• There is huge benefit in strong background knowledge before you start getting stuck into a dataset
• Data cleaning is REAL work! Hours spent just validating the structure of the data and making it usable
• I've started the very beginnings of my predicting changes in homelessness project that I've wanted to do for a really long time


• Wasnt' accepted to the masters program:
• This is sad but not the end of the line
• Currently sticking with the Udacity programs and seeing what I can make of that
• Will continue to review options as I go along


• How to add videos in html:
• The microsoft photo editing program has decided to allow you to make video editing changes
• Why, because it's fun :P See below!

Definitely the more you learn, the more you realize what you don't know and there is to learn, but I'm having a lot of fun! So many places to explore, so many mountains to climb.