<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Nicole Carlson on Medium]]></title>
        <description><![CDATA[Stories by Nicole Carlson on Medium]]></description>
        <link>https://medium.com/@parsing_science?source=rss-5e8977a4932f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*EySyoeOiNOz1gkAmooP-hQ.jpeg</url>
            <title>Stories by Nicole Carlson on Medium</title>
            <link>https://medium.com/@parsing_science?source=rss-5e8977a4932f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 11:39:59 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@parsing_science/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Diversity, Equity, and Inclusion Signals]]></title>
            <link>https://medium.com/@parsing_science/diversity-equity-and-inclusion-signals-2c91ec06ead6?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/2c91ec06ead6</guid>
            <category><![CDATA[company-culture]]></category>
            <category><![CDATA[job-hunting]]></category>
            <category><![CDATA[diversity-and-inclusion]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Thu, 24 Feb 2022 17:50:03 GMT</pubDate>
            <atom:updated>2022-02-24T17:50:03.441Z</atom:updated>
            <content:encoded><![CDATA[<p>As an outspoken, Diversity, Equity, and Inclusion (DEI) advocate, I often get asked what signals I look for in a company that indicate to me that they care about DEI. This post lists some signs I look for when I’m interviewing. If you’re also job searching, I hope this post helps provide some specific questions you can ask during your interviews.</p><h3>Job posts</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cwzr6kvqHvbDcxQaT3DhHg.jpeg" /></figure><p>A job post is often the first place I start interacting with a company so these have a huge impact on how I view a company. In the past, I have decided not to apply to companies based on language used in their job postings.</p><ul><li>Salary Bands: I will always give preference to companies that publicly list their salary bands in job postings. This is becoming even more important given that certain locations, e.g. Colorado and New York City, are going to require that salary bands be listed. I don’t want to work for a company that says they’ll just stop hiring in those locations.</li><li>Degree requirements: I am firmly against degree requirements for jobs. Some of the best data scientists I know have no degrees or non-STEM degrees. I think they can provide some signal to the hiring manager, but they should not be required.</li><li>Remote-friendly roles for everyone: Although this is less common now, historically some companies have only allowed remote work for certain roles or levels. I want to work at a company that allows everyone who can work remotely to do so. To suss this out, I look at job postings for other departments when I’m researching a company.</li></ul><h3>Interview process</h3><p>Interviews are a great place to ask questions about many of the other items in this post, but the process itself can also give you some insights into the company.</p><figure><img alt="Three giraffes “talking” to one another against a sky background." src="https://cdn-images-1.medium.com/max/1024/1*-Ksb6tsxFp5AT6RM4QFHSA.jpeg" /><figcaption>I searched “interview” and somehow this came up. Too good not to use.</figcaption></figure><ul><li>Process: The first thing I consider is whether there is a well-defined interview process that’s applied to all candidates. This lets me know that they care about evaluating everyone in the same way. However, if I’m the first data science hire, it’s possible that I will be the one putting together the process in the future.</li><li>Candidate demographics: Many companies talk about improving their interview funnels to ensure that they are interviewing and hiring folks from underrepresented groups in tech. If they aren’t collecting demographic data from all candidates, including referrals, they won’t be able to effectively measure their funnel. I know companies sometimes don’t want to scare away referred candidates, but an optional demographic survey is not too much of an ask.</li><li>Take homes: As someone who has been a hiring manager for data scientists, I understand that a take home exercise is sometimes the strongest signal into how good a candidate is. If there is a take home, I like to see if the company is respectful of my time; anything over 4–5 hours is way too much. However, not everyone can commit any time so I check to see if a company provides another option (e.g. live coding) for those who can’t. Finally in an ideal case, the company actually pays for candidates to do the take homes.</li><li>Pronouns: One small check for places that are likely to be good places to work for trans and nonbinary folks is whether the company asks for my pronouns (and uses them) and whether my interviewers introduce themselves with pronouns.</li></ul><h3>Time Off</h3><p>Given how overworked workers are in the US are, it probably comes as no surprise that how a company handles time off can tell you a lot about what they value.</p><figure><img alt="Illustration of an orange tent against a background of trees and mountains." src="https://cdn-images-1.medium.com/max/1024/1*aUbJGcWLUqv2B9Ms0X5qYg.png" /><figcaption>I haven’t camped in years, but I’d like to believe that I will camp at some point in the future.</figcaption></figure><ul><li>Vacation/PTO: the first thing I check is how much vacation they offer their employees. At this point in my career, 4 weeks is my minimum. If a company says they have unlimited PTO, I immediately ask if they measure how much PTO their employees take. Measurement is usually a good sign that the company takes PTO seriously.</li><li>Sick time: If a company offers sick time separate from vacation, they immediately get a boost in my company rankings. I hate when a company has combined buckets; it incentivizes people to work when they’re sick so they can take vacation. Often when I ask about what happens if you get sick and don’t have PTO, the company tells me, “well, we’ll make an exception if you get sick”. That is an immediate red flag to me (see <strong>Policies</strong> section below).</li><li>Parental Leave: This leave is often a good signal about whether the company cares about work/life balance. In addition to the actual amount, ask if the leave applies to both the parent giving birth and the other parent as well as what leave is available in the case of adoptions.</li><li>Gender neutral parental leave: I put this in a separate bullet point because it’s one of the things that matters the most to me. People of any gender can give birth, and not all parents identify as moms or dads. Places that still use the terms “maternity leave” or “paternity leave” immediately make me nervous. When I bring this up in interviews, sometimes a company will say that they will figure it out when the situation actually comes up, which is not a great solution.</li></ul><h3>Policies</h3><p>During my interviews, I often focus a lot on policies so I can learn more about how everyone at the company (or within a department) is treated; not just those who will be on my team.</p><ul><li>Career Ladders / Promotions: It’s a really good sign to me when a company has clearly defined career ladders and promotion processes. Although many start-ups won’t have these in place, I want to know that they’ve at least thought about these processes and considered how to prevent promotion by favoritism.</li><li>Demographics: Companies that measure their employee demographics and share them have more of my trust. If they can’t answer any questions about how many people of a specific group work at the company or were promoted, then I wonder if they are hiding something. The numbers don’t need to be perfect, but I want to work somewhere that is clearly trying to improve representation across different identities.</li><li>Policies exist: This may seem like a low bar, but companies that have set policies for various situations are better options than ones that don’t. I really don’t trust companies that tell me they do policy by exception. If their policy is that they’ll make exceptions where they need to, then they don’t actually have a policy. “Policy by exception” greatly favors outspoken people, those that have managers looking out for them, more senior people, etc. I am a big believer in working for companies that treat all employees well, not just the loudest ones.</li></ul><p>This is a non-exhaustive list of the types of things I look for in companies. I know I’m unlikely to find a place that meets everything on the list, but how they answer the questions also provides a lot of signal. I look for companies that acknowledge their current shortcomings and discuss their future plan to fix any of these issues.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2c91ec06ead6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What I’m Looking For in my Next Job]]></title>
            <link>https://medium.com/@parsing_science/what-im-looking-for-in-my-next-job-e62608625124?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/e62608625124</guid>
            <category><![CDATA[management]]></category>
            <category><![CDATA[job-hunting]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Thu, 24 Feb 2022 17:49:58 GMT</pubDate>
            <atom:updated>2022-02-24T17:49:58.240Z</atom:updated>
            <content:encoded><![CDATA[<p>This post describes what I’m looking for in my next data science leadership position. I wrote this both to help me prepare for interviews and in case it’s helpful to anyone else to see what I look for in a company. If you’re hiring and you have a role that seems like it would be a good fit for me, please send it my way!</p><p><strong>tl;dr</strong>: The most important facet for me in my next job is company culture.</p><figure><img alt="Woman in business suit pointing at a checklist" src="https://cdn-images-1.medium.com/max/1024/1*ODxmztjIXXpVi1uHacfUUw.jpeg" /><figcaption>This person creates checklists, just like me!</figcaption></figure><h3><strong>What I’m looking for in my next company</strong></h3><p>I’m pretty flexible in terms of company size, stage, domain. To me, what’s most important is working with great people, so I’m open to meeting with folks from different types of companies. My ideal company is one where there is open and transparent communication at all levels. I am a very straightforward person, and I thrive best when I have leaders I know I can be honest with.</p><h3><strong>Diversity, Equity, and Inclusion</strong></h3><p>I am someone who is extremely passionate about diversity, equity, and inclusion, so it’s important to me that I work at a company where leadership shares the same values. I believe that building an inclusive workplace needs to come from the top down, so I want to see policies and benefits that match my values. I shared some of the DEI signals I look for in companies <a href="https://medium.com/p/2c91ec06ead6">here</a>.</p><h3><strong>What I’m looking for in my next role</strong></h3><p>Within data science, my experience and interest really lies in machine learning and getting machine learning models into production. I’m looking for a job where I can continue to do some machine learning. I would love to expand my skills by managing analysts and data engineers, but I would not be ok if that was my whole job.</p><p>I believe that data scientists should be full stack where possible, including building APIs to serve up model results where appropriate or if not building APIs, then at least being heavily involved in the API process with engineers. I don’t like organizations where there’s a hand-off between data scientists and engineers.</p><h3><strong>Language</strong></h3><p>I code in Python and would like to continue working with Python data scientists, although I understand that if we’re using something like PySpark, some Scala is likely to sneak in. I am not willing to work in R.</p><figure><img alt="Python programming language logo" src="https://cdn-images-1.medium.com/max/601/1*PPIp7twJJUknfohZqtL8pQ.png" /><figcaption>Yea python!</figcaption></figure><h3><strong>What I’m looking for in my manager</strong></h3><p>I thrive best when I have a manager that I can be my authentic self with. I would like to work for someone who has experience managing large teams so I can continue growing my people management skills. It’s also important to me to work for someone who values individual career growth in addition to meeting company goals.</p><h3><strong>Location</strong></h3><p>Remote in the US. I live in Chicago, and I’m not looking to ever go back into an office. I’m a morning person so I have a slight preference for companies that schedule most of their meetings in the Eastern or Central time zones.</p><figure><img alt="Flag of the city of Chicago" src="https://cdn-images-1.medium.com/max/1024/1*1c9JDfTLv-W_qrH14NX-Xg.png" /><figcaption>Probably the best city flag ever.</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e62608625124" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What’s in a Name: SR Data Science Python Libraries 2.0]]></title>
            <link>https://medium.com/@parsing_science/whats-in-a-name-sr-data-science-python-libraries-2-0-2e87ba89a2da?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/2e87ba89a2da</guid>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Fri, 07 Aug 2020 16:12:41 GMT</pubDate>
            <atom:updated>2020-08-07T17:58:52.149Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Three hexagons with cartoon animals. The first has five octopuses, the second has a stork, and the third has a wildebeest." src="https://cdn-images-1.medium.com/max/576/1*phjBoaDSVFc27poz9k-biw.png" /><figcaption>Logos for octopod, stork, and wildebeest libraries</figcaption></figure><p>When I first started at <a href="https://shoprunner.com/">ShopRunner</a> three years ago, my coworker Hanna and I excitedly chose Harry Potter as the theme for our repo names. We both loved the series, and it seemed like it would be relatively easy to find names that loosely connected with our projects. Fast-forward to the dumpster fire known as 2020: JK Rowling (author of the Harry potter series) went into overdrive with transphobic comments on Twitter. Once I heard about those tweets, I knew we could not continue using Harry Potter names. This blog post is the story of how we tackled renaming our open source libraries.</p><p><strong>Team Decision</strong></p><p>After JK Rowling’s transphobic twitter spree, I raised the issue on Slack. I was a little nervous that not all of my team member’s would agree that this was important, but I didn’t need to worry at all. Everyone on the team immediately agreed that our names were problematic and that we should rename our repos. The only concern raised was making sure we didn’t make any changes that would break other people’s code if they were using our libraries.</p><p>We decided that the most important repos to rename were our open source projects since these repos are the public representation of our data science team. We wanted their names to be consistent with our values and to not cause harm to those who would be targeted/excluded by a naming scheme that promotes transphobia.</p><p><strong>Announcement to company</strong></p><p>At the same time that our team was discussing our repo names, our company’s LGBTQIA+ ERG was also discussing JK Rowling in their Slack group. I wanted them to know that we were taking this seriously even if there did not appear to be any immediate action.</p><p>We sent this message out to the #general Slack channel so everyone knew we were taking action:</p><blockquote>Hi all, as many of you know, for the past three years, the data science team has named all of our repos after Harry Potter related stuff. Unfortunately, JK Rowling said some terrible transphobic things this weekend.</blockquote><blockquote>Because of this, the DS team has decided we are no longer going to use Harry Potter names for our repos. We could not in good conscience call ourselves allies while using names created by JK Rowling.</blockquote><blockquote>We are in discussion within our team figuring out a new naming scheme and our process for renaming old repos. We will do our best to mitigate the risk of breaking services. Please let me know if you have any questions.</blockquote><p>We hoped that sending that message would help our company see that we were serious about putting in the work to rename our libraries.</p><figure><img alt="Yellow hexagon containing a cartoon wildebeest sitting down with the word “wildebeest” underneath" src="https://cdn-images-1.medium.com/max/360/1*56dWWovHw4Ex7GSnE0L8sQ.png" /><figcaption>Logo for wildebeest library</figcaption></figure><p><strong>Adding notes to existing repos</strong></p><p>We also wanted to get a message out to any public consumers of our libraries. Our team had three open source repos with Harry Potter affiliated names. Our first step was to remove all Harry Potter imagery from the repos and leave a version of this note (after consulting with our People team):</p><blockquote>Note: Our team previously had a tradition of naming projects with terms or characters from the Harry Potter series, but we are disappointed by J.K. Rowling’s persistent transphobic comments. In response, we will be renaming this repository, and are working to develop an inclusive solution that minimizes disruption to our users.</blockquote><p>A small step, but an important one to show we were aware of the issue and making plans.</p><p><strong>Technical Plan for deprecating the old names</strong> :</p><p>Our next step was determining how the heck you actually rename a library on <a href="http://pypi.org/">PyPI</a>.</p><p>Google was not super helpful so I turned to Twitter for help:</p><figure><img alt="Tweet from parsing_science: Hi #python friends, does anyone know what the best practices are for renaming a package on @pypi" src="https://cdn-images-1.medium.com/max/817/1*ajgMRF3VdGwkEjLsjuOKYQ.png" /><figcaption>Tweet asking for help with PyPI renaming</figcaption></figure><p>Luckily, <a href="https://twitter.com/di_codes">Dustin Ingram</a> wrote back and gave us some advice. He suggested different options depending on whether we wanted to force people to upgrade or if we were willing to cause breaking changes, etc. Side note: Dustin is one of the PyPI maintainers; <a href="https://www.youtube.com/watch?v=AQsZsgJ30AE">his PyCon 2018 talk</a> taught me that “PyPI” is pronounced pie — pee — eye, not pie — pie.</p><p>We decided our goal was to transition people away from the old library name to the new name without deleting the old versions. Dustin suggested we publish one final version of the old library with a warning about the renaming and then republish the library with the new name.</p><p><strong>Choosing a new naming scheme</strong></p><p>Now we had to choose a new naming scheme for our repos. This was by far the most difficult part of the whole process since my teammates (and I) are extremely opinionated.</p><p>After a lot of debate, we came up with the following criteria:</p><ul><li>Not taken on<a href="https://pypi.org/"> PyPI</a> (especially important for libraries if they are potential candidates for open-sourcing)</li><li>Has some kind of meaningful connection to what the repo does — e.g. <strong>canaria</strong> (Latin name for the canary species) for the model performance reporting repo because a model performance report acts as a “canary in a coal mine” for model problems</li><li>Single word (to avoid confusion around underscores vs. hyphens)</li><li>Inclusive (e.g. not associated with racist tropes)</li><li>[Optional] easy for our users to read and spell (if possible, choose words that people are likely to be familiar with)</li></ul><p>We ultimately settled on using animals as our naming scheme: either a species or specific animal. Animals seemed like a safe bet because we thought we were unlikely to discover that an animal was a jerk (except for sea lions; those pinnipeds are mean!). An added benefit was we were able to keep some of our existing repo names since they were already animals, e.g. our API named <strong>newt</strong>.</p><p>Some discarded ideas included:</p><ul><li>condiments (“what condiment means recommendation systems?”)</li><li>plants (“similar to animals but more boring”)</li><li>Chicago landmarks (“what if the architect/artist turns out to be a jerk”, sadly a highly likely outcome).</li></ul><p>The hardest part of choosing new names was finding ones that were free on PyPI. Most of the common animals were already taken so we had to start digging deeper.</p><p>However, even the animal naming scheme had some potential issues. One name we considered for our multi-task learning library was <a href="https://en.wikipedia.org/wiki/Balaur">Balaur</a> because the neural networks contain multiple task heads. A balaur is a multi-headed dragon in Romanian folklore. My partner is Romanian, and after consulting with him and his family, we realized this would be cultural appropriation. We think this demonstrates how important it is to think through repo names and ask tough questions before choosing a final version.</p><p><strong>Adding deprecation warnings</strong></p><p>One of our data scientists, <a href="https://twitter.com/paramorganic">Morgan Cundiff</a>, took the lead on the technical work. First, she created a branch where she added a warning message to every importable item in the library (https://github.com/ShopRunner/octopod/pull/18/files). This was the message we used:</p><blockquote>rename_message = ‘Tonks has been renamed to Octopod. Please install and use the new package name from PyPI.’</blockquote><figure><img alt="Hexagon containing five different colored octopuses on a light blue background with the word “octopod” underneath" src="https://cdn-images-1.medium.com/max/360/1*L650MoM-TaurX04Iz_7Cxw.png" /><figcaption>Logo for octopod library</figcaption></figure><p><strong>Renaming the repo</strong></p><p>We knew we wanted the renamed repo to be released right after the deprecation warning release so she also completed the branch where we changed the name <strong>tonks</strong> to <strong>octopod</strong> everywhere (<a href="https://github.com/ShopRunner/octopod/pull/19">https://github.com/ShopRunner/octopod/pull/19</a>).</p><p>One important part of the process was renaming the Github repo itself so we consulted with Cloud Ops to make sure someone on our team had the correct permissions on Github. We decided to rename the repo instead of creating a new one to help force our users to switch to the new name.</p><p>One deliberate choice we made was to have no breaking changes in the first release. This was important to us because we wanted to reduce the friction of having to update to a new library.</p><p>Once we merged in these branches, our rename was complete!</p><figure><img alt="Blue hexagon with a cartoon stork carrying a bundle containing the Apache Spark logo. The word “Stork” is at the bottom." src="https://cdn-images-1.medium.com/max/552/1*YsSxPuM_tRsus5yo9sqksQ.png" /><figcaption>Logo for stork library</figcaption></figure><p><strong>Introducing our new libraries: Octopod, Stork, and Wildebeest</strong></p><p>Now we are pleased to introduce you to:</p><ul><li><strong>Octopod</strong> (<a href="https://pypi.org/project/octopod/">https://pypi.org/project/octopod/</a>), formerly known as Tonks, the ShopRunner multi-task deep learning library.</li><li><strong>Stork</strong> (<a href="https://pypi.org/project/stork/">https://pypi.org/project/stork/</a>), formerly known as apparate, a tool to manage libraries in Databricks in an automated fashion.</li><li><strong>Wildebeest</strong> (<a href="https://pypi.org/project/wildebeest/">https://pypi.org/project/wildebeest/</a>), formerly known as creevey, a file processing framework, designed for IO-bound workflows that involve reading files into memory, doing some processing on their contents, and writing out the results.</li></ul><p>We hope you enjoy using our newly renamed libraries!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2e87ba89a2da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Solving D414: Section has no content]]></title>
            <link>https://medium.com/@parsing_science/solving-d414-section-has-no-content-940d88e422da?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/940d88e422da</guid>
            <category><![CDATA[numpy]]></category>
            <category><![CDATA[linting]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[docstring]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Tue, 07 May 2019 20:25:22 GMT</pubDate>
            <atom:updated>2019-05-07T20:25:22.107Z</atom:updated>
            <content:encoded><![CDATA[<p>Today I went down a long rabbit hole trying to solve a linting error for a coworker. Our team at <a href="https://shoprunner.com/">ShopRunner</a> uses <a href="https://numpydoc.readthedocs.io/en/latest/format.html#id4">numpy docstrings</a>, and we check our code with <a href="https://pypi.org/project/flake8/">flake8</a> and<a href="https://pypi.org/project/flake8-docstrings/"> flake8-docstrings</a>.</p><p>My coworker had written a method that looked like this:</p><pre>def my_sweet_method(self, examples):</pre><pre>   &quot;&quot;&quot;<br>   This is a dope docstring.<br>   <br>   Parameters<br>   ----------<br>   examples<br>       str inputs to be tokenized</pre><pre>   Returns<br>   -------<br>   Tokenized strings<br>   &quot;&quot;&quot;<br>   ...</pre><p>When we ran flake8, we kept getting this error “D414 Section has no content” on the first line of the docstring (“This is a dope docstring”). This error was super confusing to us since the docstring clearly has content.</p><p>I spent about an hour trying to research this D414 error and found nothing useful online. My problem-solving method ended up being me deleting various parts of the docstring and running the linter over and over again.</p><p>Eventually, I realized that the error seemed to be related to the Parameters section of the docstring, which was clearly not blank. After a much longer time than it should have taken, I finally got the error to disappear when I replaced the real content under Parameters with random text.</p><p>Why did this fix the error?</p><p>Well, it turns out that flake8 does not look for sections via --------; instead it does keyword matching to find specific sections as designated by numpy. Examples is one of the special keywords that is often a section in numpy docstrings. Because we had named one of the variables examples and documented it in the docstring, the linter thought there were three sections: Parameters, Examples, and Returns.</p><p>So lesson learned: do not name variables after numpy sections.</p><p>For a full list of possible sections, see here: <a href="https://numpydoc.readthedocs.io/en/latest/format.html#sections">https://numpydoc.readthedocs.io/en/latest/format.html#sections</a>. There are probably more than you think.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=940d88e422da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[One week at the Recurse Center]]></title>
            <link>https://medium.com/@parsing_science/one-week-at-the-recurse-center-431439e7dddd?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/431439e7dddd</guid>
            <category><![CDATA[recurse-center]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Wed, 17 Apr 2019 14:43:28 GMT</pubDate>
            <atom:updated>2019-04-17T15:22:48.720Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nQXI4wc__XfbW5iP2gQNDQ.jpeg" /><figcaption>Cow (“borrowed” from a coworker) hanging out in Recurse Center library with the “Oh Shit, Git” zine</figcaption></figure><p><strong>Before RC</strong></p><p>I first heard about the <a href="https://www.recurse.com/">Recurse Center</a> in 2016. RC hosts self-directed educational retreats for batches of programmers; basically go hang out in Brooklyn with cool people who like learning (yes, please!). I knew I would love to attend one of their retreats, but it was not feasible for me to go to NYC for 6 or 12 weeks (how could I possibly leave my dog for that long!?). In 2018, however, RC started doing <a href="https://www.recurse.com/blog/127-a-new-way-to-join-the-rc-community">one week mini retreats</a>, which seemed liked a more realistic/workable amount of time, but the cost of one week in NYC was still a hurdle.</p><p>Luckily, in December 2018, RC posted about <a href="https://www.recurse.com/blog/145-fellowships-for-women-working-on-open-source-programming-projects-research-and-art">fellowships for women working on open source projects</a>. This was perfect! I had an <a href="https://pypi.org/project/pymc3-models/">open source library</a> that I wanted to work on, and the fellowship would cover my living costs.</p><p>The process from application to acceptance into the program was really quick. I applied on Saturday Dec. 8, had a faculty interview with Alicia (one of the RC facilitators) on Tuesday Dec. 11, had a pairing interview with an alum on Wed Dec. 12, and heard I was accepted on Friday Dec. 14. My batch ran from Jan. 7–11, so it was a bit of a whirlwind experience. Thankfully, the application process was very smooth and well-described on their website.</p><p>I am very fortunate that my company <a href="https://shoprunner.com/">ShopRunner</a> was willing to give me a week off of work to attend RC. ShopRunner is very supportive of personal development, which is one of my favorite things about working here!</p><p><strong>At RC</strong></p><p>I’ve known quite a few recursers, but I still wasn’t sure what to expect on the first day. I was nervous about being around such a motivated group of strangers (honestly, any group of strangers would be scary.). Thankfully, I found out that RC would host a gender diversity and mini batch dinner on Sunday Jan. 6. Not only was the subject of the dinner important to me (as many of my coworkers know), but it would also allow me to meet some folks in an informal setting the day before the batch began. I knew having some familiar faces on the first day of the retreat would help make me more comfortable.</p><p>The first day was overwhelming. There were about 80 people at RC when I attended, which was apparently one of the biggest groups ever. We did tours of the space and every single person introduced themselves briefly. We also did a number of random pairings so we could meet some folks one-on-one. As an introvert, I found it fun, but also exhausting.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Q3_Vz6xKs2Olbl3aahHAfQ.jpeg" /><figcaption>Cow (“borrowed” from coworker) hanging out on RC couches with blankets and cross-stitches on the wall</figcaption></figure><p>Once the batch actually started, I had to start thinking about how I would structure my work for the next few days. The first couple of days, I felt like I needed to make a ton of progress on my library, <a href="https://github.com/parsing-science/pymc3_models">pymc3_models</a>, to justify RC giving me a spot and ShopRunner for giving me time off of work. I spent hours on the couches of the quiet floor of RC working on my library. I did try to use this library as an opportunity to collaborate with other recursers, but unfortunately, when I posted about my library to try and see if anyone wanted to collaborate, I did not receive replies.</p><p>Some things I accomplished:</p><ul><li>Got PyPI to render my description correctly (this took me over a year!)</li><li>Added a linter to pymc3_models</li><li>Added Travis CI to pymc3_models to automate linting and unittests</li><li>Reviewed two very large Pull Requests from two collaborators</li><li>Merged one of the large Pull Requests in</li><li>Added a Logistic Regression model</li><li>Various other style/bug fixes</li></ul><p>As the week went on, however, I tried to spend more time on the social floor to interact with folks. This proved to be fruitful as I got to do a bunch of random fun things.</p><p>Some things I did with other people:</p><ul><li>Made multiple Mobius Strips and experimented with cutting them</li><li>Introduced a group of recursers to “everybody is it” tag</li><li>Lost spectacularly at Ticket to Ride on board game night</li><li>Participated in a Feelings Check-In (which was amazing!)</li></ul><p>While I enjoyed being able to socialize with recursers in various activities, in practice I felt like I didn’t belong with the other folks at RC. Most people seemed quite intellectual and willing to discuss technical topics at the drop of a hat. My brain tends to work best when I intersperse really technical work with breaks to discuss more light-hearted topics. I also generally need to know someone on a more personal level before discussing really intellectual things. Luckily, I was able to find some folks who seemed to operate a bit more like me, and was even able to organize a climbing outing to Brooklyn Boulders one night with a few other climbers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T7NR0k0U3k6o_JNRyEEqZA.jpeg" /><figcaption>Cow (“borrowed” from coworker) on top of a robot built by another recurser</figcaption></figure><p><strong>After RC</strong></p><p>For me, RC was important because I got the mental space to come back to my library. I had a lot of anxiety around my library because I had been dealing with some difficult contributors and very long PRs. I had actually left my library for six months because I needed space away from all the stress. Without RC, I would not have had a dedicated week and safe space to get over my anxiety and come back to pymc3_models.</p><p>I think I would have gotten more comfortable at RC if I had more than one week to spend (I also lost almost an entire day to a bedbug infestation, which didn’t help my time management; go NYC!). There’s only so much that can happen when you’re trying to both meet people and do work. From what I can tell, finding the balance between networking and getting things on your to-do list done is tricky for a lot of people who attend RC.</p><p>I regret not participating in the pair programming workshop and working more directly with other people. I know I made a lot of progress on my library, but I wonder if my time would have been better spent collaborating. The beauty of RC though is that once you attend, it’s very possible to come back for another batch!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=431439e7dddd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Spend ¾ of Your Yearly Budget in 3 Weeks: A PySpark Cautionary Tale]]></title>
            <link>https://medium.com/shoprunner/how-to-spend-%C2%BE-of-your-yearly-budget-in-3-weeks-a-pyspark-cautionary-tale-2e91d12f9d6f?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/2e91d12f9d6f</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Mon, 04 Mar 2019 16:32:31 GMT</pubDate>
            <atom:updated>2019-12-11T21:30:06.690Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/300/0*6uGSI2iXrPG2tgKH.jpg" /></figure><p>This is a cautionary tale of how I spent a large amount of money running PySpark clusters and what I discovered about deploying a recommendation engine at scale. I’ll describe how I finally chased down an intermittent error and lessons learned along the way including why you can’t always write PySpark code like Python code and how important it is to check third party library integrations.</p><p><em>Originally published at </em><a href="https://shoprunnerblog.wordpress.com/2019/03/04/how-to-spend-%c2%be-of-your-yearly-budget-in-3-weeks-a-pyspark-cautionary-tale/"><em>https://shoprunnerblog.wordpress.com</em></a><em> on March 4, 2019.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2e91d12f9d6f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/shoprunner/how-to-spend-%C2%BE-of-your-yearly-budget-in-3-weeks-a-pyspark-cautionary-tale-2e91d12f9d6f">How to Spend ¾ of Your Yearly Budget in 3 Weeks: A PySpark Cautionary Tale</a> was originally published in <a href="https://medium.com/shoprunner">ShopRunner</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[PyData NYC 2018: How to spend 3/4 of your budget in 3 weeks]]></title>
            <link>https://medium.com/shoprunner/pydata-nyc-2018-how-to-spend-3-4-of-your-budget-in-3-weeks-ffe567327558?source=rss-5e8977a4932f------2</link>
            <guid isPermaLink="false">https://medium.com/p/ffe567327558</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[shoprunner]]></category>
            <category><![CDATA[pyspark]]></category>
            <dc:creator><![CDATA[Nicole Carlson]]></dc:creator>
            <pubDate>Thu, 14 Feb 2019 22:42:34 GMT</pubDate>
            <atom:updated>2019-02-14T22:42:34.935Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/648/1*2BNf54iXlzno9LwYeK3uTQ.png" /></figure><p>This past October, I traveled to NYC to give a talk about spending a lot of ShopRunner’s money.</p><p>Here’s the video: <a href="https://www.youtube.com/watch?v=SFkmrgfwEyE">https://www.youtube.com/watch?v=SFkmrgfwEyE</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ffe567327558" width="1" height="1" alt=""><hr><p><a href="https://medium.com/shoprunner/pydata-nyc-2018-how-to-spend-3-4-of-your-budget-in-3-weeks-ffe567327558">PyData NYC 2018: How to spend 3/4 of your budget in 3 weeks</a> was originally published in <a href="https://medium.com/shoprunner">ShopRunner</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>