![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*D2N470BiLX9Opvzndt6d-A.png\")
<\/figure>Comet<\/a> is an experimentation platform that allows you to keep track of your machine learning experiment. Another intriguing fact about Comet is that we can use it to perform exploratory data analysis. Comet integrates with popular Python visualization libraries which can help us achieve our EDA goals.<\/p>\n We will learn how to integrate Seaborn<\/strong> and Matplotlib<\/strong> with Comet in this tutorial. We will accomplish this by running some EDA on an online e-commerce dataset and logging the visualization to the Comet experimentation website or platform. Without further ado let\u2019s get started.<\/p>\n So, to begin our journey, we must ensure that the following libraries are present on our machines:<\/p>\n If you don\u2019t have the Comet library installed on your machine you can do that by typing the following command on your command prompt.<\/p>\n \u2014 or \u2014<\/p>\n Now that those libraries have been put in place. The next thing will be to obtain the dataset. So the dataset I will be using in this tutorial is from UCL and can be downloaded here<\/a>. This is a dataset that contains information about transactions that occurred in an online retail store in the United Kingdom between December 1, 2009, and December 9, 2011.<\/p>\n Let\u2019s import the necessary libraries and load the dataset into a DataFrame.<\/p>\n After you\u2019ve loaded your data into a dataframe. The next step will be to perform exploratory data analysis on it.<\/p>\n EDA is essential because it provides insight into the nature of your data. You will gain all of the necessary insights from your data using EDA. EDA informs you if your data is missing or if there are any inconsistent values in your data so that you can process it later. We\u2019ll do some EDA in this tutorial to make sure the article isn\u2019t too bogus.<\/p>\n Let us now look at the countries that purchased the most from the store during the time period. We\u2019ll take a look at the top 10 and visualize the information. In this part, we will be using Matplotlib.<\/p>\n The next EDA we might want to look at is which months had the highest number of sales over the years. To achieve this we can do so by looking at the distribution of the Invoice date column.<\/p>\n After we\u2019ve completed our visualization, the next thing will be to log the visualization to the Comet platform. In this tutorial, we will create a new project using Comet\u2019s experiment library.<\/p>\n To proceed with this phase, you will need an API key from Comet. If you haven\u2019t already signed up for their platform, click here<\/a> to do so. You can find your API key by clicking on your profile picture, then going to the settings icon, and finally scrolling down, as shown below.<\/p>\n Once that\u2019s done we can now create our experiment and log the visualization. To achieve that we can do that by typing this whole code.<\/p>\n\n
pip install comet_ml<\/span><\/pre>\nconda install -c comet_ml<\/span><\/pre>\nLoading Dataset<\/h2>\n
import pandas as pd\nimport seaborn as sns\nimport matplotlib.pyplot as plt\n\ndf = pd.read_excel(\"online_retail_II.xlsx\")\n\ndf.head()<\/pre>\n
![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*lLuTtAPjvNKe6Hl79ByP_Q.png\")
<\/figure>Exploratory Data Analysis (EDA)<\/h1>\n
mask_df = df[\"Country\"].value_counts().head(10)\nfig1 = plt.figure(figsize=(12, 10))\nplt.bar(x = mask_df.index, height=mask_df, color=\"sienna\")\nplt.xlabel(\"Country\")\nplt.ylabel(\"Counts\")\nplt.title(\"Numbers of Orders from Countries Over the Years\")\nplt.xticks(rotation=45);<\/pre>\n
![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*m5L5z5LHvFZxXipgontTEQ.png\")
<\/figure>fig2 = plt.figure(figsize=(12, 10))\nsns.histplot(df[\"InvoiceDate\"], color=\"darkslategrey\", bins=50)\nplt.title(\"Distribution of the Invoice Date\");<\/pre>\n
![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*O9t1sZQwkPyAdPYhsr436g.png\")
<\/figure>Logging the Viz to the Comet Platform<\/h1>\n
![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*D6HKWxTLnqmW5tkI79XJBw.png\")
<\/figure># import comet-ml at the top of your file\nfrom comet_ml import Experiment\n\n# Create an experiment with your api key\nexperiment = Experiment(\n api_key=\"YOUR API KEY XXXX\",\n project_name=\"Matplotlib Demo\",\n workspace=\"ibrahim-ogunbiyi\",\n)\n\nexperiment.log_figure(figure_name=\"Matplotlib Viz\", figure=fig1)\nexperiment.log_figure(figure_name= \"Seaborn Viz\", figure=fig2)<\/pre>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n
![\"\"](\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/1*V2OnXT6SNOoxknWSOQYyDA.png\")
<\/figure>