Puzzled by Python? Easy ways to get crypto data.

I have all manner of questions about crypto and patterns related to crypto. I, for one, think that I normally see the biggest drops of the market over Sunday night going into Monday. I like to have data that backs up my conjectures. By the way - I am not a programmer (although "I play one on tv".) 

So when I learned that Coindesk has an API to get at data, I thought that might be a great place to start. You can even set up a web page or service that uses the data "as long as each page or app that uses it includes the text “Powered by CoinDesk”, linking to our price page."

 The instructions on Coinbase's API page looks simple enough - send some JSON, get some JSON. But I don't know how to send some JSON. I know how to write JSON. I know how to cut and paste from their sample JSON and modify it. But I don't know how to "transmit" the JSON to the API.

Help in getting the data

So I found a website that says they can help - and they offer a tutorial and sample code on using Python or R to access the data.

Now I am not a programmer (although IPOoTV), but I know how to take code and run it. And there's a lovely bit of code ready to copy and paste into a Python tool. I have seen many people speak about Jupyter Notebook as a friendly way to work with Python. There's lots of ways to get Jupyter Notebook; one of the easiest is by downloading the Anaconda Suite. Very simple to download, install, and then power up Jupyter Notebook.

Full disclosure - you can see how much I am not a programmer because of the number of iterations it took me to get the indentation right. But Jupyter Notebook led me through that easily enough, and next thing you know, I had a CSV with very useful data for analytics.

What can you do with the data?

Once you have data, there are many tools to work with it - from Excel (four years worth of daily pairs doesn't strain Excel too much) to visualization tools like Power BI or Tableau, or analysis tools like Knime or Alteryx.

The data extract of BTC to USD gives the date; the tools can convert that simply to the day of the week and then let you do your analysis. Now there are many ways to try to estimate and recreate "the change between when you went to sleep Sunday night and when you wake up Monday morning". The image shows one simple analysis: Sunday indeed had the least "increase" of all the days; Monday had a much greater average increase. Thursday showed the most variability.

Of course, YMMV.