Context A Mexican team created a model that tries to predict the right times to sell or buy crypto. The model is based on some technical analysis of the exchange movements, mainly tracking Bollinger bands and relative strength index indicators. The model also used twitter feeds and sentiment analysis to input an extra feature to the model. Content The dataset contains hourly exchange information of the following cryptocurrencies: - BAT - BCH - BTC - ETH - GNT - LTC - MANA - TUSD - XRP Each of these cryptos is then listed with their values in Mexican pesos (MXN) as well as Bitcoin (BTC). Of course, Bitcoin is not part of the cryptos that is compared against Bitcoin ?? Hourly information includes: 1. Volume 2. High 3. Low 4. Last 5. Ask 6. Bid Depending on the file, the values are in MXN or BTC, so, for example, the btc_mxn file lists the value of Bitcoin in Mexican pesos. Acknowledgements Shout out to Rafa Olvera, who first had the idea of using the same technical analysis he had used for his stock pickings in the context of crypto (albeit with some modifications).