How can someone interested in Bitcoin glean stories from the data?
And where can one find useful data? The first steps towards this will be explained in the article.
A motto closely associated with the crypto scene: Don’t trust, verify. Don’t just trust, verify. An attitude that is reflected in the abbreviation DYOR (Do your own research), which is also well-known in the Bitcoin Revolution ecosystem and after all, it is not simply about disclosure, but about making every BTC-ECHO reader an empowered crypto-citizen. Therefore, two of the data sources used will be presented in this article. For today, we will focus on Bitcoin.
A word up front: I use the programming language R for the data analysis. If you want to learn more about this powerful statistical programming language, you can watch this video. However, all the data sources mentioned in the rest of this video are also accessible with other tools such as Python.
TradingView: Excellent source for raw data
TradingView is well known far beyond the Bitcoin scene. The price analyses on our site are basically carried out using this powerful platform. TradingView is a fantastic platform for creating charts. What few know: A first step to juggling data can be exporting the TradingView data. Thus, TradingView has a quite powerful programming language, Pine, which we introduced in this article.
But what if you want to combine external data with the TradingView data? For this purpose, the platform has been offering the possibility of data export for a few months now. One click makes it possible to export both the Bitcoin rate and all indicators in the layout as a csv file:
Exporting data from TradingView made easy.
CSV files are tables in which all values of a row are separated by commas. This gives you an easy-to-read format that can also be imported into Excel or similar.
The analyst-to-be can now work with this data. In R he would be able to read it in with a simple command:
data <- read.csv(„path/to/filename.csv“).
Of course, there are other possibilities. Many fans of R swear by Tidyverse, a group of libraries for R that makes working with data extremely easy. The Tidyverse is also reflected in particular in the Bitcoin Reports. However, it is beyond the scope of this article to go into more detail.
Juggling with bitcoin price data from Coingecko
One disadvantage of the above approach is that processes cannot be automated. Why work with programming languages if you still have to click three times to process the data further? Here, a more direct access to the courses would be helpful.
For this, the analyst of Bitcoin & Co. is often not simply concerned with prices. The development of supply and market capitalisation is always the focus of the analysis. It is true that TradingView can also help here and the amount of available data is growing. But analysts are dependent on additional data sources in this respect.
In many places, working with programming interfaces, so-called APIs, is a good solution. API stands for Application Programming Interface and, simply put, enables access to data from the command line or from a programme. Coingecko is an example of such a platform with an API. The good thing is: Currently, you don’t have to register for it.
The problem: the format is a little unusual at first. Unlike a CSV file, this so-called JSON format is not a simple table. Nevertheless, it is quite easy to use in R:
The variable data will now contain three tables, each containing the price, the market capitalisation and the volume. Via data$prices, data$market_caps and data$total_volumes, these individual tables can be further processed.
Anno 1970: The Unix time format
For daten$prices, the first five rows look like this:
The raw data from an API call regarding Bitcoin.
The second column is undoubtedly the (former) Bitcoin rate, but what is the first? It is a computer-readable time, the Unix epoch. It counts from 1.1.1970 at zero in the morning in manageable time units. For this, seconds or milliseconds are often counted, as in this case.