Build dashboards within minutes, not weeks

Subscribe to our newsletter and never miss any upcoming articles

Ever since we launched, our ambition at dstack.ai has been to make the work of data scientists easily accessible to the rest of the organization so that businesses can make quick data-driven decisions. Keeping this into perspective, over the past few months, we have launched several features that allow data scientists to instantaneously publish, track revisions, share, and get feedback on data reports.

The dashboard feature is one of them. Last week, we launched the feature allowing users to build dashboards on top of existing data visualizations and datasets that have been pushed into the dstack.ai frontend.

Having said that, there are several mature products such as Power BI, Dash, Shiny, Tableau in the market that offer dashboards as a feature. One might wonder — what is the need for us to introduce the dashboard feature in the market?

We believe that there is still a vacuum in the market when it comes to operationalizing the work of data scientists into organizations. In particular, data scientists explore very specific situational questions and find new patterns in the data that might lead to important insights. For us, dstack dashboards are a means of transporting those insights from the desk of data scientists to the rest of the organization especially the non-technical audience. At the same time, dstack dashboards remove the necessity to learn complex web development skills that are typically needed to built dashboards.

In this blog post, we will show why and how the dashboard feature from dstack should be used. In doing so, we will also be building a dashboard. The dashboard built in this blog post can be viewed at https://dstack.ai/riwaj/d/7c8cfbe8-05d3-4382-a307-978ce7da3c67. We would especially like to thank Joy Gracia Harjanto as our dashboard is based on the author’s analysis. The analysis can be read at https://towardsdatascience.com/analyzing-stocks-using-r-550be7f5f20d.

Operationalize the work of data scientists

The dashboard feature in dstack takes into consideration the previous work done by data scientists. This is taken care of by two important design aspects used at dstack.ai.

  1. Data scientists use Python or R to publish their data visualizations or datasets into the dstack frontend. We call them “stacks” at dstack.ai. As a tool for data scientists, the dstack APIs are compatible and integrated with the most popular open-source libraries for data visualization and datasets in Python and R such as Matplotlib, ggplot2, Plotly, and Bokeh.

Image for post

Various open-source libraries supported by dstack APIs

If you are new to dstack.ai, you can start by setting up a profile. This requires you to create an account at https://dstack.ai. Once you have an account, you can log in and copy your token from the settings. The token and your dstack username are used to configure your profile via the command-line utility. You can use pip or conda / CRAN to access the dstack package and a command-line tool.

a. As a quick guide, let us assume you are a Python user and you want to publish a static chart. In this case, you import a method _pushframe from dstack python library.

from dstack import push_frame

Once you plot your data (in this case, Matplotlib is used as a plotting library), you can use another method _pushframe from dstack library to publish the visualization.

import matplotlib.pyplot as plt
from dstack import push_frame

fig = plt.figure()
plt.plot([1, 2, 3, 4], [1, 4, 9, 16])

push_frame("simple", fig, "My first plot")

You can learn more about using other visualization libraries at docs.dstack.ai

b. As a quick guide, if you are an R user, here is a piece of a snippet that pushes a static chart to dstack.

library(ggplot2)
library(dstack)

df <- data.frame(x = c(1, 2, 3, 4), y = c(1, 4, 9, 16))
image <- ggplot(data = df, aes(x = x, y = y)) + geom_line()

push_frame("simple", image, "My first plot")

The published charts can be accessed via the following URL: https://dstack.ai/<user>/simple.

2. Anyone who has access to the stacks that are published in dstack frontend by data scientists can make use of the feature to build dashboards. We will illustrate this with the help of an example in the next section. At any given moment, if the stack displays a single visualization that corresponds to the selected parameters, the Dashboard then displays multiple visualizations at once — also based on selected parameters.

If you have not published any stacks at dstack.ai but would like to do so, we recommend you to go through the documentation — docs.dstack.ai.

Build dashboards within a few minutes

In order to build a dashboard at dstack.ai, you as a frontend user

  • do not require web development skills such as HTML, CSS,
  • do not require application development and deployment knowhow

All you need is access to stacks that are published into the dstack frontend. In this blog post, we will use two stacks that are published at dstack and will use them to make a dashboard.

In essence, the dashboard can be built in the following three steps.

  1. Choose the stack that visualizes the stock information data for five tech companies, namely AAPL, GOOG, FB, AMZN, TSLA during the time period of January 1, 2020, to May 15, 2020.

https://dstack.ai/riwaj/stock/Covid_Share_Plots

Image for post

Source code adapted from https://towardsdatascience.com/analyzing-stocks-using-r-550be7f5f20d

2. Choose the stack that represents a dataset denoting price simulation using a random walk and Monte Carlo method for 4 trading years for the above-mentioned tech companies.

https://dstack.ai/riwaj/stock/Price_Prediction

Image for post

Source code adapted from https://towardsdatascience.com/analyzing-stocks-using-r-550be7f5f20d

3. Build the dashboard by simply adding the stacks.

Image for post

You can also change the layout of the dashboard

4. You can give a human-readable name to the dashboard. This also applies to the individual stacks that are part of the dashboard.

Image for post

Here is another example of a dashboard build using dstack.ai https://dstack.ai/otipita/d/2710328b-393a-456a-907a-0595761af188

Manage permissions and share them inside or outside the organization

Just like the stacks, dashboards in dstack are public by default but can be made private via settings.

![Image for post](miro.medium.com/max/2100/1*eFGH8cAIJDl9E1JO..

Once the dashboard is prepared, now you can now share the stack with your peers and clients — using URL, username, or sending an invitation via an email address.

Image for post

As a user, you can see both the self-created dashboards and the ones shared to you by others if you are logged in dstack frontend. If you have multiple dashboards, you can use the search functionality to find the right one.

Image for post

Notice that the number of dashboards you have access to is denoted next to ‘My Dashboards’

Engage other parts of the organization including the non-technical audience

By offering a quick way to build dashboards and an easy way to share it within and outside the organization, dstack dashboards can bring data insights to other parts of the organization other than the data science team. We feel that dstack dashboards

  • Reduces the time taken to collaborate between data scientists and other parts of the organization by allowing data scientists to instantaneously share, publish, and get feedback on dashboards with the rest of the organization.
  • Allows non-technical audience in the organization due to the nature of the interactivity that can be built on dashboards by sending multiple parameters.

Try Now —the dashboard feature is also offered for free as part of the free-tier service!

You can build up to 3 dashboards as part of the free-tier service. You can quickly use dstack.ai from your notebooks, scripts, jobs, or even applications.

Do you think we are missing out on some use cases and features? Here is our product roadmap where you can vote for features that you would like to see in the future.

https://trello.com/b/CJOnEjrr/public-roadmap

Please sign up and come back to us with feedback and suggestions. We are curious to hear your thoughts.

Thank you very much.

Feedback survey: https://forms.gle/6MSLAGaHJFvvpB5Q6

Sign up: https://dstack.ai/auth/signup

Learn more https://dstack.ai

Documentation: docs.dstack.ai

Email for feedback: team@dstack.ai

No Comments Yet