AWS Cloud and Unsupervised Learning

Clustering Crypto

Cryptocurrencies coins by Worldspectrum | Free License

Background

Generated analysis of cryptocurrencies are available on the trading market and how they can be grouped using classification. To do this I used unsupervivsed learning and Amazon SageMaker by clustering cryptocurrencies and creating plots to present results.

Main Tasks:

• Data Preprocessing: Prepare data for dimension reduction with PCA and clustering using K-Means.

• Reducing Data Dimensions Using PCA: Reduce data dimension using the PCA algorithm from sklearn.

• Clustering Cryptocurrencies Using K-Means: Predict clusters using the cryptocurrencies data using the KMeans algorithm from sklearn.

• Visualizing Results: Create some plots and data tables to present your results.

• Optional Challenge: Deploy notebook to Amazon SageMaker.

---

Data Preprocessing

Load the information about cryptocurrencies and perform data preprocessing tasks.

1. Using the CSV file, created a Path object and read the file data directly into a DataFrame named crypto_df using pd.read_csv().

2. Using the requests library, retreive the necessary data from the following API endpoint from CryptoCompare - https://min-api.cryptocompare.com/data/all/coinlist.

With the data loaded into a Pandas DataFrame, continue with the following data preprocessing tasks.

3. Keep only the necessary columns: 'CoinName','Algorithm','IsTrading','ProofType','TotalCoinsMined','TotalCoinSupply'

4. Keep only the cryptocurrencies that are trading.

5. Keep only the cryptocurrencies with a working algorithm.

6. Remove the IsTrading column.

7. Remove all cryptocurrencies with at least one null value.

8. Remove all cryptocurrencies that have no coins mined.

9. Drop all rows where there are 'N/A' text values.

10. Store the names of all cryptocurrencies in a DataFrame named coins_name, use the crypto_df.index as the index for this new DataFrame.

11. Remove the CoinName column.

12. Create dummy variables for all the text features, and store the resulting data in a DataFrame named X.

13. Use the StandardScaler from sklearn to standardize all the data of the X DataFrame.

Reducing Data Dimensions Using PCA

Used the PCA algorithm from sklearn to reduce the dimensions of the X DataFrame down to three principal components.

After reducing the data dimensions, created a DataFrame named pcs_df using as columns names "PC 1", "PC 2" and "PC 3"; used the crypto_df.index as the index for this new DataFrame.

Final DataFrame looks like:

Clustering Cryptocurrencies Using K-Means

Used the KMeans algorithm from sklearn to cluster the cryptocurrencies using the PCA data.

Performed the following tasks:

1. Created an Elbow Curve to find the best value for k using the pcs_df DataFrame.

2. After defining the best value for k, ran the Kmeans algorithm to predict the k clusters for the cryptocurrencies data. Used the pcs_df to run the KMeans algorithm.

3. Created a new DataFrame named clustered_df, that includes the following columns "Algorithm", "ProofType", "TotalCoinsMined", "TotalCoinSupply", "PC 1", "PC 2", "PC 3", "CoinName", "Class". Maintained the index of the crypto_df DataFrames as is shown bellow.

   

Visualizing Results

Created some data visualization to present the final results. Performed the following tasks:

1. Created a 3D-Scatter using Plotly Express to plot the clusters using the clustered_df DataFrame. Included the following parameters on the plot: hover_name="CoinName" and hover_data=["Algorithm"] to show this additional info on each data point.

2. Used hvplot.table to create a data table with all the current tradable cryptocurrencies. The table has the following columns: "CoinName", "Algorithm", "ProofType", "TotalCoinSupply", "TotalCoinsMined", "Class"

3. Created a scatter plot using hvplot.scatter, to present the clustered data about cryptocurrencies having x="TotalCoinsMined" and y="TotalCoinSupply" to contrast the number of available coins versus the total number of mined coins. Used the hover_cols=["CoinName"] parameter to include the cryptocurrency name on each data point.

Additional Work

Uploaded the Jupyter notebook to Amazon SageMaker and deployed it.

The hvplot and Plotly Express libraries are not included in the built-in anaconda environments, used the altair library instead.

Performed the following tasks:

1. Uploaded the Jupyter notebook and renamed it as crypto_clustering_sm.ipynb

2. Selected the conda_python3 environment.

3. Installed the altair library by running the following code before the initial imports.

!pip install -U altair

4. Useed the altair scatter plot to create the Elbow Curve.

5. Used the altair scatter plot, instead of the 3D-Scatter from Plotly Express, to visualize the clusters. Since this is a 2D-Scatter, use x="PC 1" and y="PC 2" for the axes, and added the following columns as tool tips: "CoinName", "Algorithm", "TotalCoinsMined", "TotalCoinSupply".

6. Used the altair scatter plot to visualize the tradable cryptocurrencies using x="TotalCoinsMined" and y="TotalCoinSupply" for the axes.

7. Showed the table of current tradable cryptocurrencies using the display() command.