<p>The world stage of health crisis has undoubtedly generated that all people in their daily lives find themselves in the need to increase the use of technological platforms and tools; it is in this context, where I decided to further expand my knowledge regarding the use of technological tools, since this undoubtedly contributes to facilitating the daily tasks that are carried out on a daily basis.</p>
<p>So when I learned about Marki, I loved the idea and did not hesitate to join. It is a nonprofit association that seeks to contribute to the digitalization of microenterprises through advice provided by volunteers. By belonging to Marki, I was able to easily learn about digital marketing; The most remarkable thing about this experience was the dynamics that existed in the learning process, because not only did I learn from expert mentors, but I also reinforced what I had learned with other fellow volunteers, then managed to advise another micro-entrepreneur myself.
Additionally, one of the benefits of belonging to Marki was that I got a “DataCamp Donates” membership, which consisted of accessing one year of free access to DataCamp’s courses, considered to be one of the best platforms for learning data science in the world. When I found out about this, I asked Marki&#39;s administrative team without hesitation to have free access to the DataCamp, as it would prove quite useful for those who, like me, are convinced of the use of digital tools and data analysis to achieve business growth over time.</p>
<p>So my advice would be: if you’re like me and you find yourself interested in specializing more in the use of business intelligence tools and in the areas of analysis and data science; do not hesitate to take advantage of the DataCamp membership that Marki can offer you to continue growing professionally.</p>
<p>So far, I have developed R and Python courses related to the Data Analyst Career Track. This is helping me to develop the statistical part of my thesis project; whose theme addresses the study of the socioeconomic determinants of crime in Peru. In that sense, in my thesis I have put into practice what I learned about Ron Datacamp, in aspects such as database management (use of data from the ENAHO), operationalization of variables, and analysis of correlations and regressions; this has allowed me to develop a deep analysis of my data and obtain favorable results to the hypotheses of my research.</p>
<p>In a similar manner, I am now confident enough that in the medium term I will be able to  form groups that will allow me to participate in hackathons that seek to solve social problems.</p>
<p>For now, I have participated in the first “Seedbed” of 2021, carried out by the organization Data Science Research Perú. In that seedbed I developed a project that sought the implementation of a supervised classification model to perform a Sentiment Analysis based on the techniques of Natural Language Processing (NLP) and was able to gauge approval rates for the educational program Beca 18.</p>
<p>The results were favorable, and my project was one of the eight finalists. I was able to achieve this result in part because of the courses that I took on DataCamp. First of all, the project was developed using Python, and in my case everything I know about Python I learned from DataCamp. On the other hand, the courses that I took related to the topic of Sentiment Analysis on DataCamp allowed me to understand fundamental concepts such as data preprocessing, the Bag of Words matrix, TF-IDF transformation and the use of different supervised classification models.</p>
<p>Finally, for everything mentioned in this article, I’m very thankful for the opportunity that Marki and DataCamp gave me to improve my data analysis skills. I can&#39;t wait to see what comes next!</p>
<p><p><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/c_scale,w_418/v1627404961/marki-peru-sergio-abanto-1_mbcuzb.jpg" /></a></center><p></p>
<p><em>Sergio Abanto is an economist who is ready to collaborate in projects that contribute to achieving greater welfare and social development of the people and families of Peru.</em></p>
<p><p><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1627577029/marki-peru-2_pwshdd.png" /></a></center><p></p>
<p><em>Marki Perú is a nonprofit organization that seeks to promote digitizing micro and small businesses in Peru through education. With the help of digital marketing and e-commerce experts, we train university students who volunteer to assess micro-entrepreneurs in their path of digitizing their businesses.</em></p>


https://www.datacamp.com/community/blog/marki-peru

A DataCamp Donates learner from Peru shares his journey developing his data analysis skills to achieve greater social welfare in his country.

Marki Perú and DataCamp Team Up to Digitize Micro-entrepreneurs

<h3 id="many-machine-learning-projects-fail">Many Machine Learning Projects Fail</h3>
<p>Many organizations are investing heavily in AI and machine learning. However, that does not mean organizations are reaping the value of machine learning within their organizations. According to <a target="_blank" href="https://venturebeat.com/2019/07/19/why-do-87-of-data-science-projects-never-make-it-into-production/">VentureBeat</a>, 87% of data science projects do not make it to production. In their <a target="_blank" href="https://www.youtube.com/watch?v=EzmTZlho-EI">keynote at VentureBeat’s Transform 2019</a>, Deborah Leff, CTO for Data Science and AI at IBM, and Chris Chap, Senior Vice President of Data and Analytics at Gap, argue that many teams lack the right leadership and support for conditions that lead to success in these projects.</p>
<p>Leff and Chap describe data quality, data democratization, and organizational flow issues as key hurdles that many organizations do not have the proper tools or structure to overcome. Data teams that cannot access high-quality and relevant data or work in the correct collaborative environment are not able to succeed in these projects, regardless of their competence. They also describe issues with model decay where some organizations implement models that stop working without noticing for over three years. This was especially clear with the onset of <a target="_blank" href="https://tdwi.org/articles/2021/03/08/adv-all-covid-19-pandemic-accelerating-need-for-model-monitoring.aspx">Covid-19</a>, where lockdowns forced completely different behaviors onto consumers and many retail analytics models broke in production.</p>
<p>Another important consideration in ML projects is the rise in the use of open source technologies for testing and producing models. These include open-source programming languages like Python, and its host of libraries like Scikit-learn, pandas, PySpark, and more. Changing the version of one of these in the production environment can greatly impact the success of the project or make it stop working entirely. Managing these versions and open source frameworks requires an organized system to be effective at scale.</p>
<h3 id="the-need-for-mlops">The need for MLOps</h3>
<p>Machine Learning Model Operationalization Management <a target="_blank" href="https://www.datacamp.com/community/podcast/operationalizing-machine-learning-with-mlops">(MLOps)</a> is a way of standardizing and streamlining a machine learning project’s lifecycle management. In their book, <a target="_blank" href="https://content.dataiku.com/oreilly-mlops">Dataiku</a> discusses three key factors why MLOps is becoming so important to enterprises: there are many dependencies, organizations lack data literacy and common data language, and data scientists don’t have engineering skills.</p>
<p><strong>Many dependencies:</strong> Data is a reflection of real-world behaviors and processes, and naturally changes over time (e.g., toilet paper buying habits during Covid-19). As data changes over time, which happens constantly, models in production must continuously adapt to these changes so that organizations keep solving the original problem they intended to solve with these models.</p>
<p><strong>No common data language:</strong> In the machine learning life cycle, projects involve people from a diverse set of teams with different backgrounds. This includes data scientists, data engineers, but also functional subject matter experts. These groups do not share the same skills or data fluency to communicate about data in the same way, and their alignment is crucial to deploy models in production and evaluate their success.</p>
<p><strong>Data scientists do not have engineering skills:</strong> Data scientists typically specialize in building models and evaluating them, not in writing applications to be deployed. As organizations rely on more models and more complex models, data scientists are facing a trade-off between developing and iterating on models and maintaining them in production.</p>
<h3 id="why-mlops-is-useful">Why MLOps is Useful</h3>
<p>Machine learning models are being increasingly deployed within business processes. This increase in volume necessitates the use of MLOps to mitigate risks associated with these models, deploy these models responsibly, and use them at scale.</p>
<p>There are many risks associated with deploying a model:</p>
<ul>
<li><p><strong>The model can be unavailable for some time.</strong> Technology can potentially be unreliable in the production environment. If an organization relies heavily on these models, it can face significant challenges when the technology becomes unavailable.</p>
</li>
<li><p><strong>The model can return poor results for a sample.</strong> It may perform very well for most groups but can give poor results for one group. This needs to be corrected or understood to avoid making poor business decisions based on model outputs for the underperforming group. This poor performance can also lead to ethical issues related to AI models underperforming for some protected variables (such as race, age, gender, etc.)</p>
</li>
<li><p><strong>The model can become less fair or accurate in the future.</strong> As the world changes, the data that was used in training becomes less representative of the world the model is attempting to understand. This can lead to poor performance on future data.</p>
</li>
<li><p><strong>Data science talent may leave the organization over time.</strong> Maintaining these models requires technical skills and bandwidth data scientists may not always have to ensure perfect reliability of these models.</p>
</li>
</ul>
<p>These risks scale depending on the impact the model has on the organization and the probability of one of these risk events occurring. Many of these risks can be mitigated by ensuring the training data is a good reflection of data that will be seen in the production environment and ensuring the production environment is stable as models are very sensitive to changes. MLOps is increasingly necessary for mitigating risks associated with the increasing number of machine learning models being deployed for important decisions.</p>
<p>As the use of machine learning models in decision-making becomes more prevalent, it is increasingly important to ensure they are <a target="_blank" href="https://www.datacamp.com/community/blog/the-case-for-responsible-ai">deployed responsibly</a>. There are a lot of causes of <a target="_blank" href="https://www.datacamp.com/resources/whitepapers/data-literacy-for-responsible-ai">algorithmic bias</a> that need to be avoided when decisions are automated through AI systems. Failure to successfully minimize bias can lead to punishment through <a target="_blank" href="https://www.theverge.com/2021/4/20/22393873/ftc-ai-machine-learning-race-gender-bias-legal-violation">government regulation</a> or <a target="_blank" href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3654943">loss of demand and profits</a> for a brand. To responsibly deploy AI systems, models must be trained on compliant and unbiased data sources, have interpretable and explainable results, and accountability throughout the organization where it is easy to find where in the development pipeline something potentially went wrong to fix it effectively. To deploy AI models responsibly, organizations require a streamlined ML model life cycle.    </p>
<p>Deploying more and more models requires MLOps to be successful. It is necessary to have a streamlined process to keep track of versioning, to understand model performance versus retrained model performance, and to ensure that a model continues to perform well over time. Without these MLOps, it is very difficult to successfully add value and business impact with these models.  </p>
<h3 id="the-roles-of-an-mlops-program">The Roles of an MLOps Program</h3>
<p>While MLOps is nascent and is still evolving rapidly, it is necessary to have strong people in an organization in the following roles: Subject Matter Experts, Data Scientists, Data Engineers, and Software and DevOps Engineers.</p>
<p><strong>Subject matter experts</strong> are business-oriented individuals who understand and inform how to measure and continually evaluate the success of models during training and after deployment. For a successful MLOps program, they must develop an easy-to-understand business performance measurement and a feedback loop for distinguishing models that are not performing relative to their business expectations. Additionally, <strong>risk subject matter experts</strong> are an important part of the MLOps program. Their goal is to minimize and clarify the risk a company faces due to the deployment of ML models in production.</p>
<p><strong>Course Recommendations: <a target="_blank" href="https://www.datacamp.com/tracks/data-literacy-fundamentals">Data Literacy Skill Track</a></strong></p>
<p><strong>Data scientists</strong> are responsible for building the models which address the business questions generated by the subject matter experts. These models should be deployable in production environments with production data. Data scientists should ensure these models are easily and safely deployable in production, have tests for quality assurance and potential improvements, and have easy visibility of all deployed models from one location.</p>
<p><strong>Course Recommendations: Machine Learning Scientists Career Track (<a target="_blank" href="https://www.datacamp.com/tracks/machine-learning-scientist-with-r">R</a>/<a target="_blank" href="https://www.datacamp.com/tracks/machine-learning-scientist-with-python">Python</a>), <a target="_blank" href="https://www.datacamp.com/courses/image-processing-in-python">Image Processing in Python</a>, <a target="_blank" href="https://www.datacamp.com/courses/machine-learning-with-pyspark">Machine Learning with PySpark</a></strong></p>
<p><strong>Data engineers</strong> are responsible for creating data pipelines for these models. For a successful MLOps program, performance for all deployed models must be easily visible as well as full detail for each pipeline to see where addressable issues may exist.</p>
<p>Finally, <strong>software and DevOps engineers</strong> are in charge of helping with the successful integration of these models with the other non-machine learning applications. For a successful MLOps practice, they must integrate data engineering best practices to ensure that these models can work in parallel with existing applications reliably, continuously, and seamlessly.</p>
<p><strong>Course Recommendations: <a target="_blank" href="https://www.datacamp.com/tracks/data-engineer-with-python">Data Engineer with Python Career Track</a></strong></p>
<p>There are a lot of key roles that must be successfully utilized for a successful MLOps program. Successful deployment of hundreds of models, at scale, requires members outside of the data experts, including subject matter experts and software engineers, to generate positive business outcomes. Organizations striving to develop successful MLOps programs must aim to build a common data language across all stakeholders through upskilling and continuous learning.</p>
<p><em>Learn more about <a target="_blank" href="https://www.datacamp.com/groups/business">DataCamp for Business</a>:</em> </p>
<p><em>DataCamp for Business provides an interactive learning platform for companies that need to upskill and reskill their people on data skills. With topics ranging from data literacy, and data science to data engineering and machine learning, over 1,600 companies trust DataCamp for Business to upskill their talent.</em></p>


https://www.datacamp.com/community/blog/the-past-present-and-future-of-mlops

MLOps is not just for data experts but for everyone from data teams to SMEs and to IT. Here's the rundown on why MLOps is important,  what problems it aims to solve, and how in this jargon-free blog post. 

The Past, Present, and Future of MLOps

<p><p><center><a target="_blank" href="https://www.datacamp.com/webinars/bridging-the-communications-gap-with-data-literacy"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1627651026/blog-bridging-the-communications-gap-with-data-literacy-1_oczfxa.png" /></a></center><p></p>
<p>The real business value of data literacy comes from empowering members of an organization to understand and extract insights from data. When data literacy is infused into the company’s culture, it can act as a powerful tool to break down analytics siloes and enable scalable data storytelling throughout the organization.</p>
<p>During this webinar, Marc will explore some of the key components of building a data literate organization, and what subject matter experts need to know to start thinking “data literate”. Here’s a breakdown of what to expect:</p>
<ul>
<li><p><strong>How to visualize data:</strong> Nowadays, companies work with huge amounts of data and analysts can easily lose track of the signal in the noise. Hence, It is vital that you are selective about what to include in your visualization. Data visualization combines skills from art and science to express the right information, while making data insights easily accessible to non-technical team members. To construct clear and efficient data visualization plots, it takes time, patience, and the ability to ask the right questions. Before building your plots, you should always start by asking the right questions about your data, such as: “What is the purpose of this visualization?”.</p>
</li>
<li><p><strong>Understand data types and data flows:</strong> Knowing what data types you are dealing with is vital in order to select the right visualization method. Having a good understanding about the wide variety of data types can help you perform better exploratory data analysis (EDA) and building adequate data models.</p>
</li>
<li><p><strong>Extracting insights from data</strong>: Many organizations generate large amounts of raw data; however, they fail to effectively harness its business applications.  Hence, building a culture where most team members are empowered to understand and differentiate data insights is crucial in today’s world. Understanding different data types and choosing the appropriate chart or visualization method greatly contributes to convey the message of your data insights. During this webinar Marc Wintjen will share different techniques that will empower your team to extract actionable insights from data.</p>
</li>
</ul>
<p>If you want to stay ahead of the competition, extracting actionable insights from your data is a necessity, not an option. Therefore, building a data literate culture that empowers your team to generate, capture, and analyse data will give your business a great competitive advantage over the rest of the market.</p>
<p>To learn more about how data culture can be adopted within your organization and empower your team to make better data-driven decisions <a target="_blank" href="https://www.datacamp.com/webinars/bridging-the-communications-gap-with-data-literacy">join our webinar with Marc Wintjen, Bloomberg’s Risk Analytics Architect.</a></p>


https://www.datacamp.com/community/blog/bridging-the-communications-gap-with-data-literacy

Marc Wintjen, Bloomberg’s Risk Analytics Architect and author of Practical Data Analysis using Jupyter Notebooks, will join us this 5th of August to discuss how the key principles to building data literacy in your organization can empower your team to make better and informed decisions.

Marc Wintjen, Bloomberg’s Risk Analytics Architect and author of Practical Data Analysis using Jupyter Notebooks, will join us to discuss how key principles to building data literacy in your organization can empower your team.

Bridging the Communications Gap with Data Literacy

<p>If you want to learn more about how AI Singapore is building an AI literate nation you can watch full webinar <a href="https://www.datacamp.com/resources/webinars/developing-an-ai-literate-nation">here</a> the or read the event recap <a href="https://www.datacamp.com/community/blog/event-recap-developing-an-ai-literate-nation">blog post</a>.</p>
<h3 id="how-ai-singapore-is-developing-an-ai-fluent-nation">How AI Singapore is Developing an AI Fluent Nation</h3>
<p><p><center><a target="_blank" href="https://res.cloudinary.com/dyd911kmh/image/upload/v1627569487/_ai_singapore_ma__infographics__1__snvmzr.pdf"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1627569487/_ai_singapore_ma__infographics__1__snvmzr.png" /></a></center><p></p>
<p><center><a target="_blank" href="https://www.datacamp.com/business/demo"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></p></p>


https://www.datacamp.com/community/blog/how-ai-singapore-is-developing-an-ai-fluent-nation

AI Singapore has helped grow AI talent at a national scale through their national upskilling program. Learn how they leveraged DataCamp to help Singapore become a data fluent nation

How AI Singapore is Developing an AI Fluent Nation [Infographic]

<p><p><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1626879891/employee-spotlight-maarten-1_fyzuft.png" /></a></center><p></p>
<h3 id="what-do-you-do-at-datacamp-and-what-does-a-typical-day-look-like-">What do you do at DataCamp and what does a typical day look like?</h3>
<p>I was initially a content developer in the core curriculum team, working on R and Python courses. Then I switched over to curriculum expansion, where I focus on Tableau and Power BI. It’s hard to say what a typical day looks like since it can be quite varied, but there are two ways it generally goes.</p>
<p>One is, I research what it takes to get certified in certain software, and then I put together a rough list to see if it can be made into a viable course. Then it’s a matter of creating conceptual lessons, exercises, and demo videos. The other process is where we work with an external subject matter expert and that person provides the material that should be in a course and how we should present it. Then, we implement that material into Teach, DataCamp’s tool for building courses. Currently, I’m working on two courses simultaneously, but we are improving the process so we can handle more courses at once.</p>
<h3 id="do-you-ever-teach-the-courses-yourself-">Do you ever teach the courses yourself?</h3>
<p>What I really like as a content developer is that we have that opportunity to be the instructors as well. I really like how much of the world you reach, and for me, it was really nice the first time I released a course and got feedback from people outside my network.</p>
<h3 id="where-did-your-interest-in-what-you-do-begin-">Where did your interest in what you do begin?</h3>
<p>I have a background in science, doing a PhD in aquatic ecology, and it was during that time that I came into contact with complex datasets and I was required to do my own data analysis. These were often quite specialized, and back then, R was not where it is today. I remember creating graphs in R was a pain, and I recall our first classes in it were kind of dreadful—no one really understood what we were doing. It was later in my PhD that I understood the value of being able to code your own analyses in your own ways.</p>
<p>One of the main focuses of my PhD was to investigate the effects of land use on water quality and on the presence of mosquitoes in small water bodies. After finishing my PhD, I had the opportunity to participate in a Hackathon that was centered around the spread of dengue fever, a vector-borne disease spread by mosquitoes. It was through this that I got exposed to new technologies and tools in data science, and winning an award during the Hackathon led me to my first job as a data science consultant.</p>
<h3 id="what-brought-you-to-datacamp-">What brought you to DataCamp?</h3>
<p>Since experienced data scientists were in short supply at the time, we were looking to train ourselves. And that is when I found out about DataCamp. I really liked the interactive way of teaching and the practice exercises. As a consultant, I worked on projects involving machine learning, data visualization, and creating dashboards, but my favorite part of my job was when I was giving workshops to clients or students. I thought to myself that there should be a way to do the teaching side of things as a full-time job. Also, I had a Master’s in teaching so I have always been drawn towards education, and when I saw the job opening at DataCamp I saw this as an ideal opportunity to follow that passion.</p>
<h3 id="what-advice-would-you-give-to-anyone-applying-for-a-role-at-datacamp-">What advice would you give to anyone applying for a role at DataCamp?</h3>
<p>I think a lot depends on the department they’re applying to, but in my case, I think it’s great when you know the platform, when you’ve taken some courses, and know how the underlying information is presented. Something new hires will experience is the independence that comes with working on your own projects, although by all means, help is available when needed.
Another feature of working at DataCamp is that there’s always room for improvement. New features are coming out all the time, so don’t be afraid to suggest and try new things. The content team also has varied backgrounds. In addition to myself, we have people from business engineering and economics, marketing analysis, computer science, biostatistics, psychology, geology, and sociology, and this diversity results in interesting ideas for datasets, exercises, and conceptual insights. The short of it is, it does not matter what your background is if you want to get into data science or want to start as a content developer.</p>
<h3 id="outside-of-work-you-founded-an-orchestra-called-conforte-symphonic-https-www-conforte-be-is-that-still-active-">Outside of work you founded an orchestra called <a target="_blank" href="https://www.conforte.be/">“ConForte Symphonic”</a>, is that still active?</h3>
<p>Yes, my wife and I founded it! We’ve always played in orchestras, but there are very few opportunities to play as an amateur. Therefore, we decided to set up our own orchestra late 2019. We played three charity concerts, and were making plans for the future, but then the pandemic happened. Recently, we have the green light to start playing again, and can look forward to future concerts.</p>
<p>Meanwhile, I’ve been looking into ways to use ConForte’s data to make better decisions. For example, I was able to use regression to predict the number of tickets sold for our charity concerts, and we made the right decision to play three concerts instead of our initially planned two—all three of them were sold out.</p>
<h3 id="do-you-find-there-s-a-lot-of-support-in-datacamp-for-projects-like-this-">Do you find there’s a lot of support in DataCamp for projects like this?</h3>
<p>During my onboarding, I was talking to our COO, Martijn, about the founding of our orchestra, and he was really interested to know about that. In general he’s keen to know about peoples’ interests, especially in the case of founding something like a side-project. I also see it as a win-win situation: you learn skills in your spare time that you can apply to your daily job, and vice-versa, so it really keeps you motivated in both areas.</p>
<h3 id="what-do-you-like-doing-in-your-spare-time-">What do you like doing in your spare time?</h3>
<p>I really like the fact that we are playing music again, but other than that most of my time is spent with our two (soon three) kids. As someone with a passion for teaching it’s really been nice to see them wandering about the world and taking it all in for the first time. I really like building up their understanding of things, so if they ask a question I’ll sometimes answer, other times I will ask a question in response so that they think about it—actually it’s a bit like machine learning, you could say, they’re learning by example and repetition, and that’s how machine learning works. By seeing lots of examples, they know how to repeat certain things in certain situations. Anyway, now that they’re off for the summer it’s going to be nice to be spending more time with the family.</p>
<p><em>Would you like to join Maarten at DataCamp? We’re currently hiring in New York; London; Leuven, Belgium; and remote locations. <a target="_blank" href="https://www.datacamp.com/jobs/">Check out our open roles.</a></em></p>


https://www.datacamp.com/community/blog/employee-spotlight-maarten

From mosquitoes to machine learning and music, this week’s Employee Spotlight focuses on Maarten. A former biology researcher and data science consultant, he now works at DataCamp as a content developer.

Employee Spotlight: Maarten

https://www.datacamp.com/community/blog/operationalizing-machine-learning-with-mlops

In this episode of DataFramed, Adel speaks with Alessya Visnjic, CEO and co-founder of WhyLabs, an AI Observability company on a mission to build the interface between AI and human operators. 

Operationalizing Machine Learning with MLOps

<p>Sometimes it’s good to go back to basics, so we put three of our instructors up to the task of role-playing as students getting the Tableau Desktop Specialist certification through Tableau’s official service—using what they learned when creating our <a target="_blank" href="https://www.datacamp.com/tracks/tableau-fundamentals">Tableau skill track</a> and our new course <a target="_blank" href="https://www.datacamp.com/courses/statistical-techniques-in-tableau"><em>Statistical Techniques in Tableau</em></a>. Now that they’ve come out the other side, we decided to interview them to find out about their experiences!</p>
<p><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1626876158/tableau-student-to-certification-1_mhyg40.png" /></a></center><p>


<h3>How did DataCamp help you prepare for the Tableau certification?</h3>
<p>Hadrien: You can be asked to create a visualization or asked which feature to use. DataCamp’s courses give you an overview of what’s available and what you can create with Tableau. They also dive into specific topics, which are the topics the certification is centered around—including analyzing data, creating dashboards, and connecting to data sources.</p>

<p>Sara: DataCamp’s courses cover all the topics that are necessary for the exam. The videos give you an introduction and tell you what is possible using the tool—so, it’s shown in the videos how to connect to a dataset or create a visualization. You can take that knowledge and practice it yourself in the exercises. This is really helpful because during the exam you will be asked to create these visualizations over and over again. So the more you have practiced the quicker you will be at doing these when the time comes. Which is great because it’s timed, and this will free up time for the other questions. Also, exercises are practised in a virtual machine so you get practise with a Windows interface, and this is exactly how you will be working during the exam.</p>


<h3>What surprised you most about the exam?</h3>
<p>Hadrien: Most of the questions don’t require you to build a visualization in Tableau to answer them since they are about where to find certain features—of course, some do require you to build a visualization to gather insights and select the correct answer, but this is in the minority. Also you might get asked about topics that are not in the exam preparation guide, topics which were meant to be reserved for more advanced certification levels, so there won’t be many but it’s good to be aware of that.</p>

<p>Sara: Probably that the exam was multiple choice, as opposed to more active exercises like creating visualizations, and many were related to where you could find certain features. You could also use Google when you weren’t entirely sure, so that was quite a nice surprise.</p>


<h3>What’s the best advice you’d give to someone preparing to take the exam?</h3>

<p>Sara: I’d say two things: on the one hand, practice, practice, practice, make sure you use Tableau a lot, so do things like connecting to datasets and building visualizations until you become comfortable with it. And on the other hand I’d say take a look at some practice exams online to get a feel for what will be asked, and also to identify and shore up any weak points.</p>

<h3>What can we look forward to on the new Tableau course on DataCamp?</h3>

<p>Maarten: You can expect to learn how statistical techniques work, and then see how they are implemented in Tableau. This course is meant for people who have had experience in working with Tableau but want to know more about when and how they should use certain statistical techniques, such as adding a trendline or perform clustering, and also who want to know more about what’s happening behind the scenes.</p>

<h3>What’s your favourite report type that you teach in this course?</h3>

<p>Maarten: I like the predictive reporting this course teaches, so for example you will see our virtual machine predicting the type of dinosaur based on bow length or predict the number of views a YouTube video will get.</p>


<p><center><a target="_blank" href="https://www.datacamp.com/tracks/tableau-fundamentals"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1626876551/tableau-student-to-certification-2_gsth2z.png" /></a></center></p>


https://www.datacamp.com/community/blog/tableau-student-to-certification

We talked to three DataCamp instructors who recently took the Tableau Desktop Specialist Certification and asked them about how well our courses prepared them for this.

Tableau: From Student to Certification

<section>
<p>Even though the <code>data.frame</code> object is one of the core objects to hold data in R, you&#39;ll find that it&#39;s not really efficient when you&#39;re working with time series data. You&#39;ll find yourself wanting a more flexible time series class in R that offers a variety of methods to manipulate your data.</p>

<p>&nbsp;<code>xts</code> &nbsp;or the Extensible Time Series&nbsp;is one of such packages that offers such a time series object. It&#39;s a powerful R package&nbsp;that&nbsp;provides an extensible time series class, enabling uniform handling of many R time series classes by extending <code>zoo</code>, which is the package that is the creator for an S3 class of indexed totally ordered observations which includes irregular time series.</p>

<p><code><tt>xts</tt></code> has a lot to offer to make your time series analysis fast and mistake free, but it can take some time to get used to it.&nbsp;</p>

<p>This <code>xts</code> cheat sheet provides you not only with an overview of the <code>xts</code> object, how to create and inspect them, but also goes deeper into how you can manipulate time series with <code>xts</code>: you&#39;ll see how to replace and update values, how to select, index and subset your objects, how to handle missing values and how to perform arithmetic operations.</p>

<p>Click on the button below to see (and download) the xts cheat sheet:&nbsp;</p>

<p><a href="http://datacamp-community-prod.s3.amazonaws.com/72771032-0653-4d87-a798-4a83879e99c8" target="_blank"><img alt="" src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1625838439/xts_Cheat_Sheet-_Time_Series_in_R_nbiik0.png" style="width: 800px; height: 551px;" /></a></p>

<p><strong>(Click above to download a printable version or read the online version below.)&nbsp;</strong></p>

<p>This cheat sheet will help you to get yourself up to speed in no time!&nbsp;</p>

<div id="scoped-content">
<style type="text/css">:target:before { content:""; display:block; height:150px; margin:-150px 0 0; } h3 {font-weight:normal; } h4 { 	font-weight: lighter; }
</style>
<h2>R For Data Science Cheat Sheet: xts</h2>

<p>eXtensible Time Series (xts) is a powerful package that provides an extensible time series class, enabling uniform handling of many R time series classes by extending zoo.</p>

<p>Load the package as follows:</p>

<pre>
<code>library(xts)</code></pre>

<h3>Xts Objects</h3>

<p>xts objects have three main components:</p>

<ul>
	<li>coredata: always a matrix for xts objects</li>
	<li>index: vector of any <code>Date</code>, <code>POSIXct</code>, <code>chron</code>, <code>yearmon</code>, <code>yearqtr</code>, or <code>DateTime</code> classes</li>
	<li>xtsAttributes: arbitrary attributes</li>
</ul>

<h2>Creating xts Objects</h2>

<pre>
<code>xts1 &lt;- xts(x=1:10, order.by=Sys.Date()-1:10)
data &lt;- rnorm(5)
dates &lt;- seq(as.Date("2017-05-01"), length=5, by="days")
xts2 &lt;- xts(x=data, order.by=dates)
xts3 &lt;- xts(x=rnorm(10), order.by=as.POSIXct(Sys.Date()+1:10), born=as.POSIXct("1899-05-08")
xts4 &lt;- xts(x=1:10, order.by=Sys.Date()+1:10)</code></pre>

<h3>Convert To And From xts</h3>

<pre>
<code>data(AirPassengers)
xts5 &lt;- as.xts(AirPassengers)</code></pre>

<h3>Import From Files</h3>

<pre>
<code>dat &lt;- read.csv(tmp_file)
xts(dat, order.by=as.Date(rownames(dat),"%m/%d/%Y"))
dat_zoo &lt;- read.zoo(tmp_file, index.column=0, sep=",", format="%m/%d/%Y")
dat_zoo &lt;- read.zoo(tmp,sep=",",FUN=as.yearmon)
dat_xts &lt;- as.xts(dat_zoo)</code></pre>

<h2>Inspect Your Data</h2>

<p>Extract core data of objects</p>

<pre>
<code>core_data &lt;- coredata(xts2)</code></pre>

<p>Extract index from objects</p>

<pre>
<code>index(xts1)</code></pre>

<h3>Class Attributes</h3>

<p>Get index class</p>

<pre>
<code>indexClass(xts2)</code></pre>

<p>Replacing index class</p>

<pre>
<code>indexClass(convertIndex(xts,'POSIXct'))</code></pre>

<p>Get index class</p>

<pre>
<code>indexTZ(xts5)</code></pre>

<p>Change format of time display</p>

<pre>
<code>indexFormat(xts5) &lt;- "%Y-%m-%d"</code></pre>

<h3>Time Zones</h3>

<p>Change the time zone</p>

<pre>
<code>tzone(xts1) &lt;- "Asia/Hong_Kong"</code></pre>

<p>Extract the current time zone</p>

<pre>
<code>tzone(xts1)</code></pre>

<h3>Periods, Periodicity &amp; Timestamps</h3>

<p>Estimate frequency of observations</p>

<pre>
<code>periodicity(xts5)</code></pre>

<p>Convert xts5 to yearly OHLC</p>

<pre>
<code>to.yearly(xts5)</code></pre>

<p>Convert <code>xts3</code> to monthly OHLC</p>

<pre>
<code>to.monthly(xts3)</code></pre>

<p>Convert <code>xts5</code> to quarterly OHLC</p>

<pre>
<code>to.quarterly(xts5)</code></pre>

<p>Convert to quarterly OHLC</p>

<pre>
<code>to.period(xts5,period="quarters")</code></pre>

<p>Convert to yearly OHLC</p>

<pre>
<code>to.period(xts5,period="years")</code></pre>

<p>Count the months in <code>xts5</code></p>

<pre>
<code>nmonths(xts5)</code></pre>

<p>Count the quarters in <code>xts5</code></p>

<pre>
<code>nquarters(xts5)</code></pre>

<p>Count the years in <code>xts5</code></p>

<pre>
<code>nyears(xts5)</code></pre>

<p>Make index unique</p>

<pre>
<code>make.index.unique(xts3,eps=1e-4)</code></pre>

<p>Remove duplicate times</p>

<pre>
<code>make.index.unique(xts3,drop=TRUE)</code></pre>

<p>Round index time to the next <code>n</code> seconds</p>

<pre>
<code>align.time(xts3, n=3600)</code></pre>

<h3>Other Useful Functions</h3>

<p>Extract raw numeric index of <code><tt>xts1</tt></code></p>

<pre>
<code>.index(xts4)</code></pre>

<p>value of weekday&nbsp;in index of <code>xts3</code></p>

<pre>
<code>.indexwday(xts3)</code></pre>

<p>Value of hour in index of <code>xts3</code></p>

<pre>
<code>.indexhour(xts3)</code></pre>

<p>Extract first observation of <code>xts3</code>&nbsp;</p>

<pre>
<code>start(xts3)</code></pre>

<p>Extract last observation of <code>xts4</code>&nbsp;</p>

<pre>
<code>end(xts4)</code></pre>

<p>Display structure of <code>xts3</code>&nbsp;</p>

<pre>
<code>str(xts3) </code></pre>

<p>Extract raw numeric index of <code>xts1</code></p>

<pre>
<code>time(xts1)</code></pre>

<p>First part of <code>xts2</code></p>

<pre>
<code>head(xts2)</code></pre>

<p>Last part of <code>xts2</code>&nbsp;</p>

<pre>
<code>tail(xts2)</code></pre>

<p>&nbsp;</p>

<h2>Export <code>xts</code> Objects</h2>

<pre>
<code>data_xts &lt;- as.xts(matrix)
tmp &lt;- tempfile()
write.zoo(data_xts,sep=",",file=tmp)</code></pre>

<h2>Replace &amp; Update</h2>

<p>Replace values in <code>xts2</code> on dates with <code><tt>0</tt></code>&nbsp;</p>

<pre>
<code>xts2[dates] &lt;- 0 </code></pre>

<p>Replace dates from 1961 with <code>NA</code>&nbsp;</p>

<pre>
<code>xts5["1961"] &lt;- NA</code></pre>

<p>Replace the value at 1 specific index with <code>NA</code></p>

<pre>
<code>xts2["2016-05-02"] &lt;- NA </code></pre>

<p>&nbsp;&nbsp;</p>

<h2>Applying Functions</h2>

<p>Locate endpoints by time&nbsp;</p>

<pre>
<code>ep1 &lt;- endpoints(xts4,on="weeks",k=2)  
ep2 &lt;- endpoints(xts5,on="years") </code></pre>

<p>Calculate the yearly mean&nbsp;</p>

<pre>
<code>period.apply(xts5,INDEX=ep2,FUN=mean)</code></pre>

<p>Split <code>xts5</code> by year</p>

<pre>
<code>xts5_yearly &lt;- split(xts5,f="years")</code></pre>

<p>Create a list of yearly means&nbsp; &nbsp;</p>

<pre>
<code>lapply(xts5_yearly,FUN=mean) </code></pre>

<p>Find the last observation in each year in <code>xts5</code></p>

<pre>
<code>do.call(rbind,  lapply(split(xts5,"years"), function(w) last(w,n="1 month"))) </code></pre>

<p>Calculate cumulative annual passengers</p>

<pre>
<code>do.call(rbind,  lapply(split(xts5,"years"),  cumsum)) </code></pre>

<p>Apply standard deviation&nbsp;to rolling margins of <code>xts5</code></p>

<pre>
<code>rollapply(xts5, 3, sd)  </code></pre>

<h2>Selecting, Subsetting &amp; Indexing</h2>

<h3>Select</h3>

<pre>
<code>mar55 &lt;- xts5["1955-03"] #Get value for March 1955</code></pre>

<h3>Subset</h3>

<p>Get all data from 1954&nbsp;</p>

<pre>
<code>xts5_1954 &lt;- xts5["1954"] </code></pre>

<p>Extract data from Jan to March &lsquo;54&nbsp;</p>

<pre>
<code>xts5_janmarch &lt;- xts5["1954/1954-03"]</code></pre>

<p>Get all data until March &lsquo;54&nbsp;&nbsp;</p>

<pre>
<code>xts5_janmarch &lt;- xts5["/1954-03"] </code></pre>

<p>Subset <code>xts4</code> using <code>ep2</code>&nbsp;</p>

<pre>
<code>xts4[ep1] </code></pre>

<h3><code>first()</code> and <code>last()</code></h3>

<pre>
<code>first(xts4,'1 week') #Extract first 1 week
first(last(xts4,'1 week'),'3 days') #Get first 3 days of the last week of data</code></pre>

<h3>Indexing</h3>

<p>Extract rows with the index of <code>xts3</code>&nbsp;</p>

<pre>
<code>xts2[index(xts3)] </code></pre>

<p>Extract rows using the vector <code>days</code>&nbsp;</p>

<pre>
<code>days &lt;- c("2017-05-03","2017-05-23") 
xts3[days] </code></pre>

<p>Extract rows using <code>days</code> as <code>POSIXct</code>&nbsp;</p>

<pre>
<code>xts2[as.POSIXct(days,tz="UTC")] </code></pre>

<p>Index of weekend days</p>

<pre>
<code>index&lt;-which(.indexwday(xts1)==0|.indexwday(xts1)==6) </code></pre>

<p>Extract weekend days of <code>xts1</code>&nbsp;</p>

<pre>
<code>xts1[index] </code></pre>

<h2>Missing Values</h2>

<p>Omit NA values in <code>xts5</code>&nbsp;</p>

<pre>
<code>na.omit(xts5) </code></pre>

<p>Fill missing values in <code>xts2</code>&nbsp;using last observation</p>

<pre>
<code>xts_last &lt;- na.locf(xts2) </code></pre>

<p>Fill missing values in <code>xts2</code> using next observation</p>

<pre>
<code>xts_last &lt;- na.locf(xts2,  fromLast=TRUE)</code></pre>

<p>Interpolate <code>NA</code>s</p>

<pre>
<code>na.approx(xts2)  </code></pre>

<h2>Arithmetic Operations</h2>

<h3><code>coredata()</code> or <code>as.numeric()</code></h3>

<pre>
<code>xts3 + as.numeric(xts2) #Addition
xts3 * as.numeric(xts4) #Multiplication
coredata(xts4) - xts3 #Subtraction
coredata(xts4) / xts3 #Division</code></pre>

<h3>Shifting Index Values</h3>

<p>Period-over-period differences&nbsp;</p>

<pre>
<code>xts5 - lag(xts5) </code></pre>

<p>Lagged differences</p>

<pre>
<code>diff(xts5,lag=12,differences=1) </code></pre>

<h3>Reindexing</h3>

<pre>
<code>xts1 + merge(xts2,index(xts1),fill=0) #Addition
xts1 - merge(xts2,index(xts1),fill=na.locf) #Subtraction</code></pre>

<h2>Merging</h2>

<pre>
<code>merge(xts2,xts1,join='inner') # Inner join of xts2 and xts1
merge(xts2,xts1,join='left',fill=0) #Left join of xts2 and xts1,
rbind(xts1, xts4)</code></pre>

<h2>Going Further</h2>

<p>Want to know more about <code>xts</code>? Check out DataCamp&#39;s <a href="https://www.datacamp.com/courses/manipulating-time-series-data-in-r-with-xts-zoo">Manipulating Time Series Data in R with </a><a href="https://www.datacamp.com/courses/manipulating-time-series-data-in-r-with-xts-zoo">xts</a><a href="https://www.datacamp.com/courses/manipulating-time-series-data-in-r-with-xts-zoo">&nbsp;&amp; zoo</a>&nbsp;course!&nbsp;</p>
</div>
</section>


https://www.datacamp.com/community/blog/r-xts-cheat-sheet

Get started on time series in R with this xts cheat sheet, with code examples. 

community

datacamp

Official Blog