{"id":8530,"date":"2024-01-09T06:00:03","date_gmt":"2024-01-09T14:00:03","guid":{"rendered":"https:\/\/live-cometml.pantheonsite.io\/?p=8530"},"modified":"2025-04-24T17:03:35","modified_gmt":"2025-04-24T17:03:35","slug":"how-to-integrate-comet-with-catboost-workflows","status":"publish","type":"post","link":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\/","title":{"rendered":"How to Integrate Comet with Catboost Workflows"},"content":{"rendered":"\n
\n
\n
\n
\n
\n
\"person<\/figure>
<\/picture><\/div>\n<\/div>
Photo by Campaign Creators<\/a> on Unsplash<\/a><\/figcaption><\/figure>\n

Catboost is one of the most versatile gradient-boosting models. Its crucial capability is processing categorical data without converting it to numerical data. This means that the model can perform its function as you desire after specifying categorical data. An added benefit is that a data practitioner can easily and quickly establish a baseline with minimal data transformations. Unfortunately, as you scroll through Comet\u2019s supported libraries, you will discover the glaring lack of Catboost workflows support. Fortunately, I have a clever but straightforward workaround leveraging Comet\u2019s versatile capabilities.<\/p>\n

Comet\u2019s support for Tensorflow\u2019s Tensorboard can become a saving grace for anyone using Catboost effectively with Comet. One integral feature of Catboost is that it uses Tensorboard to keep track of training runs and stores that information locally on the machine\u2019s disk. Let\u2019s look at the general workflow of performing this integration.<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n

Requirements<\/h2>\n

There are a few things you need to install before you go through with this:
\n1. Catboost.<\/p>\n

2. Comet\u2019s official library.<\/p>\n

3. A Comet account that you can get by signing up here<\/a>.<\/p>\n

3. Visual Studio Code (or any IDE that supports Tensorboard Integration).<\/p>\n

With the above, it is now possible to proceed.<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n

Catboost Workflows Integration with Tensorboard<\/h2>\n

Catboost integrates with Tensorboard to ensure we can adequately visualize training runs. It indicates the desired metric that one has specified during the training run, graphs it, and allows for customization. The picture below is an excellent example of what I am talking about.<\/p>\n

\n
\n
\"tensorboard<\/figure>
<\/picture><\/div>\n<\/div>\n
Tensorboard for Catboost, screenshot by author<\/figcaption>\n<\/figure>\n

An added advantage of this library is that it stores the information in a local directory or a directory you initially specified before the training run, as seen below.<\/p>\n

\n
\"Catboost<\/figure>
<\/picture><\/div>\n
Catboost directory by author<\/figcaption>\n<\/figure>\n

In this article, we use this feature to our advantage as Comet allows us to upload this information and create a custom panel to have a clear view of the given data. Despite the lack of support for this library, we see that the support for TensorFlow\u2019s Tensorboard gives us an added advantage because we can leverage a single aspect of the Catboost library and upload critical information through an existing Comet function.<\/p>\n

Let\u2019s code!<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n

Simple Project<\/h2>\n

Our simple project here will only focus on the capabilities Catboost, Tensorboard, and Comet offer. All data transformation has already been done, so we have a reasonably clean dataset that will be fed into the model.<\/p>\n

The dataset of our choice is the Bengaluru House Prediction Dataset from Machinehack<\/a>. This competition aims to test your skills in regression problems. We can now check our preprocessed data with Pandas to have a general overview of our data.<\/p>\n

import<\/span> pandas as<\/span> pd\n\n#reading data in local directory<\/span>\ndf = pd.read_csv(\"preprocessed_train_data.csv\"<\/span>)\n\n#defining features(X) and targets(y)<\/span>\nX = df.drop([\"price\"<\/span>], axis=1<\/span>)\ny = df[\"price\"<\/span>]\n\n#printing the top 5 positions of X and y<\/span>\nprint<\/span>(X.head())\nprint<\/span>(y.head())<\/span><\/pre>\n
\n
\n
\"dataset\"<\/figure>
<\/picture><\/div>\n<\/div>\n
X data by author<\/figcaption>\n<\/figure>\n
\n
\"dataset\"<\/figure>
<\/picture><\/div>\n
y data by author<\/figcaption>\n<\/figure>\n
Now that we have seen our myriad of features (including one categorical one), we can incorporate this into Comet.<\/p>\n
Comet Incorporation<\/h2>\n
Note: Import the Comet library first and initialize the project before proceeding with any other code for a smoother way forward.<\/p>\n
First, import the Comet library and initialize our project under the name \u201ccatboost_comet.\u201d<\/p>\n
import<\/span> comet_ml\ncomet_ml.login(project_name=\"catboost_comet\"<\/span>)<\/span><\/pre>\n
We then define our categorical features and perform train-validation splits for the training of our model. Catboost requires a user to specify the categorical features that a dataset has.<\/p>\n
from<\/span> sklearn.model_selection import<\/span> train_test_split\nimport<\/span> numpy as<\/span> np\n\n#defining categorical features<\/span>\ncategorical_features_indices = np.where(X.dtypes !=np.float<\/span>)[0<\/span>]\n\n#Train-test split<\/span>\nX_train, X_validation, y_train, y_validation = train_test_split(X, y, train_size = 0.8<\/span>, random_state = 12<\/span>)\n<\/span><\/pre>\n
Now, we can feed the above information into Catboost\u2019s regressor and perform training.<\/p>\n
from<\/span> catboost import<\/span> CatBoostRegressor\n\n#defining model params<\/span>\nmodel  = CatBoostRegressor(iterations=50<\/span>, depth=3<\/span>, learning_rate=0.1<\/span>, loss_function='RMSE'<\/span>, early_stopping_rounds=5<\/span>)\n\n#training model on data<\/span>\nmodel.fit(X_train, y_train, cat_features=categorical_features_indices, eval_set=(X_validation, y_validation), plot=True<\/span>)\n\n#performing inference<\/span>\ny_valid = model.predict(X_validation)\n<\/span><\/pre>\n
\n
\n
\"training<\/figure>
<\/picture><\/div>\n<\/div>\n
Training run and results by author<\/figcaption>\n<\/figure>\n
After this run, there is the expectation that information on the training run will be stored in a folder for the Tensorboard to perform a visualization. This folder will be within the project\u2019s directory. We can extract two files containing the critical info we need from it.<\/p>\n
\n
\"local<\/figure>
<\/picture><\/div>\n
Local directory by author<\/figcaption>\n<\/figure>\n
Within the \u201ccatboost_info\u201d directory, we will find the \u201clearn_error.tsv\u201d and \u201ctest_error.tsv\u201d files. We shall log these files into Comet using \u201clog_table().\u201d<\/p>\n
#logging both tables to Comet<\/span>\nexperiment.log_table(\".\/src\/catboost_info\/learn_error.tsv\"<\/span>)\nexperiment.log_table(\".\/src\/catboost_info\/test_error.tsv\"<\/span>)<\/span><\/pre>\n
Once those are complete, we can end the experiment and open Comet to see whether it was successful.<\/p>\n
#Ending experiment<\/span>\nexperiment.end<\/span>()<\/span><\/pre>\n
Comet Visualization<\/h2>\n
Once we open our Comet project page, we will find and open our project.<\/p>\n
\n
\"Catboost<\/figure>
<\/picture><\/div>\n
Project image by author<\/figcaption>\n<\/figure>\n
\n
\n
\"screenshot<\/figure>
<\/picture><\/div>\n<\/div>\n
Screenshot of project by author<\/figcaption>\n<\/figure>\n
After pressing the prominent blue button in the middle, you will get the menu below and pick \u201cData Panel.\u201d<\/p>\n
\n
\n
\"Comet<\/figure>
<\/picture><\/div>\n<\/div>\n
Comet Menu by author<\/figcaption>\n<\/figure>\n
After opening that menu, we will find a drop-down menu called \u201cData Selection,\u201d where we will see the two tables we initially uploaded.<\/p>\n
\n
\n
\"adding<\/figure>
<\/picture><\/div>\n<\/div>\n
Screenshot by author<\/figcaption>\n<\/figure>\n
We then pick the first table, and it will give a preview of the data. After we press it, it will appear on the original page with no panels, as seen below.<\/p>\n
\n
\n
\"screenshot<\/figure>
<\/picture><\/div>\n<\/div>\n
Screenshot by author.<\/figcaption>\n<\/figure>\n
When we press \u201cAdd\u201d in the top right corner, we see the option to add another panel, and then we can repeat the process for the other table that we logged in.<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n
\n
\n
\n
We finally have logged information concerning Catboost workflows using Comet\u2019s support for the Tensorboard. In this tutorial, we have successfully integrated a library that Comet does not officially support.<\/p>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"
Photo by Campaign Creators on Unsplash Catboost is one of the most versatile gradient-boosting models. Its crucial capability is processing categorical data without converting it to numerical data. This means that the model can perform its function as you desire after specifying categorical data. An added benefit is that a data practitioner can easily and […]<\/p>\n","protected":false},"author":79,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"customer_name":"","customer_description":"","customer_industry":"","customer_technologies":"","customer_logo":"","footnotes":""},"categories":[23,7],"tags":[],"coauthors":[176],"class_list":["post-8530","post","type-post","status-publish","format-standard","hentry","category-integrations","category-tutorials"],"yoast_head":"\nHow to Integrate Comet with Catboost Workflows<\/title>\n<meta name=\"description\" content=\"Learn how to integrat Catboost workflows, a library that Comet does not officially support, in this step by step tutorial.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"How to Integrate Comet with Catboost Workflows\" \/>\n<meta property=\"og:description\" content=\"Learn how to integrat Catboost workflows, a library that Comet does not officially support, in this step by step tutorial.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\" \/>\n<meta property=\"og:site_name\" content=\"Comet\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/cometdotml\" \/>\n<meta property=\"article:published_time\" content=\"2024-01-09T14:00:03+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-04-24T17:03:35+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY\" \/>\n<meta name=\"author\" content=\"Mwanikii Njagi\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@Cometml\" \/>\n<meta name=\"twitter:site\" content=\"@Cometml\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Mwanikii Njagi\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"7 minutes\" \/>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"How to Integrate Comet with Catboost Workflows","description":"Learn how to integrat Catboost workflows, a library that Comet does not officially support, in this step by step tutorial.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows","og_locale":"en_US","og_type":"article","og_title":"How to Integrate Comet with Catboost Workflows","og_description":"Learn how to integrat Catboost workflows, a library that Comet does not officially support, in this step by step tutorial.","og_url":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows","og_site_name":"Comet","article_publisher":"https:\/\/www.facebook.com\/cometdotml","article_published_time":"2024-01-09T14:00:03+00:00","article_modified_time":"2025-04-24T17:03:35+00:00","og_image":[{"url":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY","type":"","width":"","height":""}],"author":"Mwanikii Njagi","twitter_card":"summary_large_image","twitter_creator":"@Cometml","twitter_site":"@Cometml","twitter_misc":{"Written by":"Mwanikii Njagi","Est. reading time":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#article","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\/"},"author":{"name":"Mwanikii Njagi","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/c7043b3e6b992af7b3220aa1f27d2162"},"headline":"How to Integrate Comet with Catboost Workflows","datePublished":"2024-01-09T14:00:03+00:00","dateModified":"2025-04-24T17:03:35+00:00","mainEntityOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\/"},"wordCount":833,"publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#primaryimage"},"thumbnailUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY","articleSection":["Integrations","Tutorials"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows\/","url":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows","name":"How to Integrate Comet with Catboost Workflows","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#primaryimage"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#primaryimage"},"thumbnailUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY","datePublished":"2024-01-09T14:00:03+00:00","dateModified":"2025-04-24T17:03:35+00:00","description":"Learn how to integrat Catboost workflows, a library that Comet does not officially support, in this step by step tutorial.","breadcrumb":{"@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#primaryimage","url":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY","contentUrl":"https:\/\/miro.medium.com\/v2\/resize:fit:700\/0*oygcvjT9V-Z9plzY"},{"@type":"BreadcrumbList","@id":"https:\/\/www.comet.com\/site\/blog\/how-to-integrate-comet-with-catboost-workflows#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.comet.com\/site\/"},{"@type":"ListItem","position":2,"name":"How to Integrate Comet with Catboost Workflows"}]},{"@type":"WebSite","@id":"https:\/\/www.comet.com\/site\/#website","url":"https:\/\/www.comet.com\/site\/","name":"Comet","description":"Build Better Models Faster","publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.comet.com\/site\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.comet.com\/site\/#organization","name":"Comet ML, Inc.","alternateName":"Comet","url":"https:\/\/www.comet.com\/site\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/","url":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","contentUrl":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","width":310,"height":310,"caption":"Comet ML, Inc."},"image":{"@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/cometdotml","https:\/\/x.com\/Cometml","https:\/\/www.youtube.com\/channel\/UCmN63HKvfXSCS-UwVwmK8Hw"]},{"@type":"Person","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/c7043b3e6b992af7b3220aa1f27d2162","name":"Mwanikii Njagi","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/image\/1a3c516cf04aca9418dfb2213081f4df","url":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2023\/08\/cropped-1_2jy9gyk0G_yaniWm8gJFVA-1-96x96.webp","contentUrl":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2023\/08\/cropped-1_2jy9gyk0G_yaniWm8gJFVA-1-96x96.webp","caption":"Mwanikii Njagi"},"url":"https:\/\/www.comet.com\/site\/blog\/author\/freddynjagigmail-com\/"}]}},"_links":{"self":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8530","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/users\/79"}],"replies":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/comments?post=8530"}],"version-history":[{"count":1,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8530\/revisions"}],"predecessor-version":[{"id":15407,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/8530\/revisions\/15407"}],"wp:attachment":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/media?parent=8530"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/categories?post=8530"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/tags?post=8530"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/coauthors?post=8530"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}