This term I am currently taking a Blockchain module. One of my assignments included analsying the performance of a crypto coin in STATA. I was assigned the coin ICX.
The bulk of the task involved data manipulation including the cleaning, sorting and merging of the data alongside running regressions in STATA. I initially needed to gather relevant datasets. Each dataset provided different insights into my coins overall performance over its life cycle.
I began to sort the data via the date column and clean the data by identifying and removing outliers as well as removing any columns that weren’t necessary for my regression analysis. I found the best way to Identify these outliers is to install the ‘extremes’ package, following this up with the command – extremes ‘column name’ , IQR(3). This command returns the extremes for the column name chosen within a 300% range of the IQR (Inter Quartile Range). The return can then be used to remove the relevant observations.
The merging was the most intense part of my project as I had to carry out 16 merges in order to obtain a full concatenated dataset with all the relevant columns I needed for my analysis. I decided to merge 1:1 as this felt the most logical approach given the format of my data. Each merge was done on the ‘Date’ Column, as I had imported each dataset from a csv I had to tsset the date columns but also there were certain dates that were presented in a weekly format rather than a daily format hence for these datasets I had to assign a new date column that would store the conversion from weekly to daily dates. See a small snippet of my code below.
Having completed the merging I was now able to regress on my data and draw conclusions based on my outputs.
I achieved 91% in this project, see below.
