<?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 Leonidas Boutsikaris on Medium]]></title>
        <description><![CDATA[Stories by Leonidas Boutsikaris on Medium]]></description>
        <link>https://medium.com/@leobouts?source=rss-9bae4e527b64------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*ucX4eft2_64A13VDzZQU_g.png</url>
            <title>Stories by Leonidas Boutsikaris on Medium</title>
            <link>https://medium.com/@leobouts?source=rss-9bae4e527b64------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 21:41:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@leobouts/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[A blog about git that will actually help you at work]]></title>
            <link>https://leobouts.medium.com/a-blog-about-git-that-will-actually-help-you-at-work-16917062c53e?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/16917062c53e</guid>
            <category><![CDATA[git]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[bitbucket]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[version-control]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Sun, 13 Feb 2022 10:06:02 GMT</pubDate>
            <atom:updated>2022-02-13T10:06:02.393Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Photo of metro lines" src="https://cdn-images-1.medium.com/max/1024/1*UzXdL73GnkC36WI5Jb1MOg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@unarchive?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Jeremy Bezanger</a> on <a href="https://unsplash.com/?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>What would we do without git? pass usbs around? I’m sure you already know what git is and that’s why you’re here so i&#39;m skipping the long intros and dive right in.</p><p>We will go through some tips and scenarios that are really common in the development lifecycle.</p><p>At first let’s stay clear about what we’re talking about. Most software companies use the following git structure with many variations but the base idea holds true. Let me draw that for you.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bWR8qtA3hcloWPBQ1uREIw.png" /></figure><p>The main branch is used only for release cycles (when your favourite app has an update), for example every week or month, and has to be stable. That means no bugs, no problems, etc. (who am I even kidding?).</p><p>The development branch is used for the between periods of these cycles. In this branch a feature is merged after its completion.</p><p>Finally we have the individual feature branches, these are the things you work on.</p><p>When you start working on something, you “branch out” (that’s the slang) from the stable development branch, you append your changes, do some testing and finally you merge your feature in the development branch. After all this fuzz everything that made it to the development branch gets merged to the main branch and is released.</p><p>A thing to not miss out is that you always have to rebase before doing a merge into the development branch. Your feature was being developed for some time (days or weeks). That means you were left behind after you branched out and missed commits that were merged into the development branch (bug fixes, features, an API you might need? etc).</p><p>Pull the remote master locally, checkout in your branch and hit this command<br> <br>&gt;$git rebase master</p><p><em>Side note: Things can go wrong, for example having some conflicts while rebasing. Resolve them with your favourite tool and then use</em></p><p>&gt;$git rebase --continue</p><p>or you can decide that this is not the day and just</p><p>&gt;$git rebase --abort</p><h3><em>Stay committed (or not)</em></h3><p>Are you sure you’re writing the right commit messages? (if not that’s okay). <br>Commit messages in feature branches are basically useful when collaborating with other engineers and also when you forget what you coded the day before. That’s why over describing your commit messages is useful for you <strong><em>only if its useful for you</em></strong>. If you know what you’re doing go ahead and just put a “changes” message.</p><p>If you actually want to make it classy, instead of using the inline message with git commit -m<strong><em>, </em></strong>drop the<strong><em> </em></strong>-m parameter and an editor will open. Write as a single line a brief description, let an empty line and get in the details there. This will come up really useful if someone actually reads it and has a nice title-paragraph format.</p><p>The same applies for how you group your commits. If you want to make it right, differentiate the changes in chunks that make sense, not just big file changes.</p><p>We all know the git addcommand. But what about its arguments? <br>You should definitely try the -p parameter.</p><p>&gt;$git add file.cpp -p</p><p>This will make Git go to each individual change you made in this file and ask you if you want to stage it. That way you can easily differentiate your commits and seperate it in pieces that make sense.</p><p><strong>But why good commit messages and splitting commits into pieces is not necessary?</strong></p><p>When your feature is ready you will try to merge it into the development branch. While merging you should probably squash your commits into one<br>meaning that if you have 10 individual commits <strong>they should become one</strong>.<br>This is needed to avoid filling the development branch with tons of commits.</p><p>You can craft one big beautiful commit message by including the number and the name of your feature ticket, a brief overview of it and a paragraph description that gets into more details about the actual changes. This actually can be done in the process of squashing the commits. Squashing is really easy.</p><ol><li>&gt;$git rebase -i HEAD~x</li></ol><p>where x are the number of the commits you have.</p><p>After hitting enter a vim or vi editor will open (insert not funny and not overplayed joke about vim here).</p><p>2. Hit the a button that stands for append and now you can interactively write in the terminal.</p><p>3. Change every pick with an s except the first one.</p><p>4. Hit the esc button and write :wq and enter</p><p>5. A new editor will open, do the same thing (presa), and delete all lines that do not start with a #.</p><p>6. Then you can finally write your beautiful descriptive commit message.</p><h4>Don’t forget your feature issue number!</h4><p>A lot of tools and developers use the commit messages to match the code to the appropriate issues and its really useful.</p><p>7. Hit the esc button and write :wq again and you’re done.</p><p>Now you have a different history from the branch that is uploaded to github/bitbucket etc. You should push using the -f parameter. This will re-write the branch history on the remote side.</p><p><strong><em>Please inform any other devs about that if they are working on this branch before doing so because they will be affected by these changes!</em></strong></p><h3>what about now, what about today?!</h3><p>After rebasing, the first commit probably has an old timestamp. You can easily change that by using:</p><p>&gt;$git commit --date=now</p><h4>OK, you did that, you have a ready squashed your commit and forgot to upload a string change.</h4><p>Git add and then commit by using the amend argument. This creates a change in the previous commit without the fuzz of rebasing again.</p><p>&gt;$git commit --amend</p><p>An editor will open, you can just :wq and exit or modify the commit message and save. Force push again because this rewrites the history as mentioned before!</p><h4>Does your colleague works on some stuff you need to use too?</h4><p>For example an API might be in the workings but not fully ready. Until this is merged to the development branch and you rebase to get it you can use the first iteration of it instead of mocking your responses! This can help you develop faster and better.</p><p>&gt;$git cherry-pick commit-hash-here</p><p>This will get all the code of this commit on top of your branch. You could also use the branch name, if you’re bored to look for the hash, and get the HEAD of the branch into yours.</p><p>&gt;$git cherry-pick branch-name-here</p><h3>Let’s say you commited something you shouldn&#39;t</h3><p>First question, did you pushed them in the remote? <br>If the answer is yes you have to do a force push at the end. If not continue with your normal flow.</p><p>You can do a git reset or a git revert to go back to a previous commit. The first one moves the pointer of the HEAD of your branch to the commit you will specify while the later will just create a new commit undo the changes there and then commit it. No strong opinions here, use what seems better to you.</p><p>&gt;$git revert HEAD~x..HEAD where x are the number of the last commits commit you want to revert. <br> <br> &gt;$git revert commit-sha1 commit-sha2 … for multiple commits also.</p><p>We can add the --soft or --hard parameter while resetting.</p><p>&gt;$git reset --soft HEAD~x this argument will preserve your changes but in the staging area. With the same fashion the --hard parameter will not keep your changes in the temporary staging area.</p><h4>Why did you changed the code authors?</h4><p>Line endings are evil, try to use the same ones as your remote repository. If you don’t whole files will be committed as one big addition. Generally speaking that’s not inherently bad. When someone though needs information about a specific chunk of code in this file they will git blame and they will see your name. Then they will message you and you will not have a clue about the inner workings of this code. You should definitely change this.</p><ol><li>Rename the file you want to restore the original authors</li><li>&gt;$git checkout master -- filepath-of-original-file-here/file.cpp <br>This will get you the original file from the master untouched.</li><li>Vimdiff new-file-path old-file-path <br>This will open a two-way splitted editor inside the terminal, like merging conflicts!</li><li>Type ]c to go to each individual change each time, then hit dp to keep the changes only. Apply to all change patches.</li><li>When you are done with it just :wq on both editors.</li><li>&gt;$rm renamed-file-path</li><li>&gt;$git add file &amp;&amp; git commit --amend (do the usual stuff we talked above)</li></ol><h4>The other way around!</h4><p>Suppose a developer starts by writing the skeleton of a class in a commit and you implement the whole thing. You then rebase in the first commit of the branch so you can merge with a single commit in the development branch and now the author of this file according to git blame seems to be the one that made the first commit. If someone tries to reach them about this chunk of code they will also not have a clue.</p><p>You can easily change that by using the --author argument.</p><p>&gt;$git commit --amend --author=”Author Name &lt;email@address.com&gt;”</p><p>Don’t forget to push with force!</p><h3>The keep your house clean section</h3><p>While developing, the project will have countless external libraries, package managers etc, these files are downloaded on compile time and do not exist on the repository. Lot’s of time when versions and packages change intermediate versions of code stay in your local branch and disturb the compiling process leaving you wondering why this damn thing is not running.</p><p>&gt;$git clean -f -d -x</p><p>will remove everything that is not on the remote repository. Now try compiling again.</p><h4>Get the trash out</h4><p>From time to time run the &gt;$git prune command. This is an internal housekeeping utility that cleans up unreachable or &quot;orphaned&quot; Git objects. Unreachable objects are those that are inaccessible by any refs. Any commit that cannot be accessed through a branch or tag is considered unreachable.</p><h4>Are you still in the same company?</h4><p>After some time you may end up having lots of local branches that are now delete on the remote side because they are merged in the development branch or even test branches from stuff you tried. These slow you down and you should throw them away.</p><pre>#remove all the local branches that are not on the remote side</pre><pre>&gt;$git fetch -p &amp;&amp; git branch -vv | grep &#39;: gone]&#39; | awk &#39;{print $1}&#39; | xargs git branch -D</pre><pre>#remove all local branches that are not on remote and that are fully merged and that are not used</pre><pre>git fetch -p &amp;&amp; git branch --merged | grep -v &#39;*&#39; | grep -v &#39;master&#39; | xargs git branch -d</pre><ul><li>&gt;$git fetch -p will prune all branches no longer existing on remote</li><li>&gt;$git branch -vv will print local branches and the pruned branch will be tagged with the gone keyword</li><li>grep &#39;: gone]&#39; selects only the branches that are gone</li><li>awk &#39;{print $1}&#39; filters the output to display only the name of the branches</li><li>xargs git branch -D will loop over all lines (branches) and force remove the branches (-D needs to be capital for branches that are fully merged)</li></ul><p><strong>Git is awesome and you should definitely master it! Can you think any other useful commands or situations? Feel free to share your thoughts.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=16917062c53e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Polarization on Social Media]]></title>
            <link>https://leobouts.medium.com/polarization-on-social-media-87bea2cea267?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/87bea2cea267</guid>
            <category><![CDATA[society]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[social-science]]></category>
            <category><![CDATA[social-media]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Wed, 03 Feb 2021 15:21:44 GMT</pubDate>
            <atom:updated>2021-02-03T15:21:44.822Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Spider web with droplets of water representing the social bubles" src="https://cdn-images-1.medium.com/max/1024/1*-plEj7ogSAPBrmen9ktOdA.jpeg" /></figure><h3>What is Polarization?</h3><p>Polarization describes the division of people into two contrasting groups or sets of opinions or beliefs. The term is used in various domains such as politics and social studies. For example political polarization refers to the divergence of political attitudes to ideological extremes. Social studies use this term to describe the segregation within a society in terms of income inequality or social and class status.</p><p>Currently social media are one the biggest sources for news and information as more and more people go online to discuss and socialize. Polarization is linked with harmful effects such as intensifying stereotypes and creating echo chambers.</p><p>In echo chambers an individual get their news only from like-minded people as they share and reinforce one another’s opinions. Additionally, the fact that people tend to ignore opposing views, in combination with algorithmic personalization, results in a significant increase of polarization.</p><h3>Social and Psychological Factors</h3><figure><img alt="A psychology proffesional taking notes" src="https://cdn-images-1.medium.com/max/1024/1*dwgL8hRDcWrmmR5_fhWeeg.jpeg" /></figure><p>Individuals experience discomfort when given data that actively challenge their opinions. In the field of psychology, cognitive dissonance occurs when a person holds two or more contradictory beliefs, ideas, or values and experiences psychological stress because of that. In simple terms dissonance is defined as a the lack of agreement.</p><p>They also want to reduce the discomfort that is caused from cognitive dissonance. Reduction occurs by strengthening opinions that come in agreement with their own and downplaying everything that challenges them. This leads individuals to a selective exposure on information . Selective exposure is also demonstrated in groups where people assign themselves with certain social identities.</p><p>We can understand this better with the self-categorization theory. This theory stems from the social identity theory, which holds that conformity stems from psychological processes. Accordingly, proponents of the self-categorization model hold that group polarization occurs because individuals identify with a particular group and conform to a prototypical group position that is more extreme than the group mean. It is shown that groups of people tend to make decisions that are more extreme than the initial inclination of its members.</p><h3>Polarization online</h3><p>Online entities such as news or social media platforms are aware of their users opinions and aim to maximize their satisfaction. As discussed above, platforms will present content in a way that minimizes psychological stress. This leads to media bias.</p><p>Media bias is the bias or perceived bias of journalists and news producers within the mass media in the selection of many events and stories that are reported and how they are covered. When this happens online, personalization of the content creates algorithmic bias.</p><p>Algorithmic bias describes systematic and repeatable errors in a computer system that create unfair outcomes, such as privileging one arbitrary group of users over others.</p><p>Bias can emerge due to many factors like the design of the algorithm. Due to personalization we don’t see the same content and this is the main reason for the formation of filter bubbles.</p><h3>Filter Bubbles</h3><p>Filter bubbles are the echo chambers of social media. In news media, an echo chamber is a metaphorical description of a situation in which beliefs and opinions are strengthened by communication and repetition inside a closed system. It is important to distinguish the difference between echo chambers and filter bubbles. This two concepts are almost identical, however, filter bubbles are a result of algorithms that choose content based on previous online behaviour, as with search histories or online shopping activity.</p><h3>Polarization and society</h3><p>Political polarization can be defined as the difference in ideological extremes but in political science almost in every context polarization is considered as the gap between the political parties of a society.</p><p>Most of the time political parties disagree on policy issues and that is the main drive of democracy. With heightened polarization the followers of each political party start fearing that the other will destroy their society with their agendas.</p><p>Destroying the other side becomes their only objective and this is how democracies fall apart.</p><p>Social networks are frequently liable for terrorism. Terrorist leaders create communities of individuals that have the same opinions and fuel them with each other. As mentioned above, when like-minded people discuss with each other they tend to move toward extreme positions. This has a bigger effect when people are already quite extreme.</p><p>Terrorist leaders know this and they try to make sure that all individuals inside this community will speak and interact with people that have the same extreme direction. If members of the community think that they have a shared identity the polarization will grow. Terrorist leaders will also repress opposing views and will not tolerate internal disagreement. They take every step needed to ensure unity.</p><p>Most individuals lack in confidence on their own views or have more confidence that are willing to show. Fear of marginalization or being proven wrong make them present a moderate version of themselves. In either case, group dynamics can push people toward a more extreme position.</p><p>Social influence also plays a great role. People have a certain image of themselves and how they want to be viewed by others. Most people like to think of themselves as not identical to but as different from others, but only in the right direction and to the right extent. There is evidence that social influence is an independent factor behind group polarization; consider in particular the fact that mere exposure to the views of others can have this effect, even without any discussion at all .</p><p>Combining these factors we obtain a highly dangerous and highly polarized community for everyday life.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=87bea2cea267" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Go serverless, Functions as a Service.]]></title>
            <link>https://leobouts.medium.com/go-serverless-functions-as-a-service-7ab3b7f0ba30?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/7ab3b7f0ba30</guid>
            <category><![CDATA[serverless]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[microservices]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Wed, 11 Nov 2020 17:05:46 GMT</pubDate>
            <atom:updated>2020-11-11T17:05:46.416Z</atom:updated>
            <content:encoded><![CDATA[<h3>Serverless?</h3><p>The serverless architecture provides services on demand. Well, servers are still used but a company is charged from the cloud vendor <strong>only on its usage</strong> and not for a fixed amount of servers. This happens by triggering them to work only when needed. These services are auto-scalling and cut a lot of daily costs and deployment time.</p><p>Nowadays we use modular separate components to build an application. This is known as the microservice architecture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/744/1*C1ya1Jo0DF9qT-1_ELXHfw.png" /><figcaption>microservices</figcaption></figure><h3>Function as a Service?</h3><p>Function as a service is a serverless way to execute modular pieces of code. This enables developers to write small pieces of code that only executes in response of a user action.</p><p>A lot of cloud companies provide serverless solutions such as Google Cloud Functions or Azure functions but limit your choices.</p><h3>OpenFaaS</h3><p>Moving to open source the <a href="https://www.openfaas.com/">OpenFaaS</a> project provides a stack and ensures that any process can become a function through a docker container. So the code can be run from whatever language you are currently writing and not tied to the cloud vendor product.</p><h3>A quick example</h3><p>For starters we need <a href="https://www.docker.com/products/container-runtime">docker</a> installed and running.</p><blockquote>Watch out, if you are using a mac with a small disk space you definitely want to go to docker settings and make the docker raw image smaller, something like 16GB would be enough. This file takes an absurd amount of “logical space”, even if it’s not actually that big it fools the OS.</blockquote><p>Also let’s have the OpenFaas cli installed</p><blockquote>$ curl -sSL <a href="https://cli.openfaas.com/">https://cli.openfaas.com</a> | sudo sh</blockquote><p>Download the git repo</p><blockquote>$ git clone <a href="https://github.com/openfaas/faas">https://github.com/openfaas/faas</a></blockquote><p>We now have to initialize a Docker Swarm. A Docker Swarm is a group of either physical or virtual machines that are running the Docker application and that have been configured to join together in a cluster. Docker swarm is a container orchestration tool, meaning that it allows the user to manage multiple containers deployed across multiple host machines.</p><blockquote>$ docker swarm init</blockquote><p>Swarm init will give you a password so you can access the localhost. You can save it with</p><blockquote>$ faas-cli login -p “password”</blockquote><p>We can now deploy the FaaS stack and go to <a href="http://localhost:8080/">http://localhost:8080</a>.</p><blockquote>$ cd faas<br>$ ./deploy_stack.sh</blockquote><h3><strong>Create a function</strong></h3><p>Create a new folder and cd inside it.</p><p>At first we can download the FaaS templates</p><blockquote>$ faas-cli template pull</blockquote><p>And check which we can use with</p><blockquote>$ faas-cli new --list</blockquote><p>Let’s create a python function</p><blockquote>$ faas-cli new new_function --lang python3</blockquote><p>This will create 3 files. The handler.py the requirements.txt and the new_function.yml</p><p>The requirements.txt file is used to install dependencies for your code. For example numpy, tensorflow or anything else you need.</p><p>Build and deploy.</p><blockquote>$ faas-cli build -f ./new_function.yml &amp;&amp; faas-cli deploy -f ./new_function.yml</blockquote><p>You can test your function either by curl or with the OpenFaaS GUI from localhost.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7ab3b7f0ba30" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Setting asynchronously an image downloaded from a url with Picasso in an android Preference.]]></title>
            <link>https://leobouts.medium.com/setting-asynchronously-an-image-downloaded-from-a-url-with-picasso-in-an-android-preference-990bb7a4d951?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/990bb7a4d951</guid>
            <category><![CDATA[android]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[android-app-development]]></category>
            <category><![CDATA[android-apps]]></category>
            <category><![CDATA[androiddev]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Wed, 11 Nov 2020 15:35:30 GMT</pubDate>
            <atom:updated>2020-11-11T15:35:30.733Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sn3UAWAYUAuTGcy9x_gEwQ.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@juleskwak?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Julie Kwak</a> on <a href="https://unsplash.com/s/photos/picasso-painting?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Synchronous requests are easy to read and debug. The thread freezes until you get a response back. You handle errors and that’s it.</p><p>The thing is that we want users to continue interacting with the application while the program waits to get a response back. The error handling and debugging becomes a bit more complicated.</p><p>In our case we can leave all this handling to <a href="https://square.github.io/picasso/">Picasso</a>, a powerful image downloading and caching library for Android.</p><p>Suppose that we have a custom preference category for the settings of my application with various sub sections.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/5aa3274cf3379b6fad43d016f8de2568/href">https://medium.com/media/5aa3274cf3379b6fad43d016f8de2568/href</a></iframe><p>Now as we can see above, the custom preference uses a custom layout that handles how the image should be in terms of sizes and margins with the ImageView. Its is important to match the android id of the custom layout.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/891770cba7fa576d6fa9cdec4d170f58/href">https://medium.com/media/891770cba7fa576d6fa9cdec4d170f58/href</a></iframe><h4><strong>Don’t forget the dependencies</strong></h4><p>Inside your build.gradle file add the following line, gradle sync and verify that you build correctly and that you can import Picasso.</p><blockquote>implementation “com.squareup.picasso:picasso:(latest version here)”</blockquote><p>The current version as of November 2020 is : 2.71828</p><p>Now inside the kotlin file that controls this fragment we can use Picasso to fill this image from a url we have.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9efd2319ef75723e7c87834fc844e80d/href">https://medium.com/media/9efd2319ef75723e7c87834fc844e80d/href</a></iframe><p>Finally we could make this picture round so it can be displayed nice and beautifully. Note that this works well for square images so you can adjust this corner radius according to your needs.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/35fe11604a1afb86c6b0e0f1e74c63ca/href">https://medium.com/media/35fe11604a1afb86c6b0e0f1e74c63ca/href</a></iframe><p>Awesome, we just updated our settings icon!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=990bb7a4d951" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Rank-1 updates in Linear Algebra in simple terms.]]></title>
            <link>https://leobouts.medium.com/rank-1-updates-in-linear-algebra-in-simple-terms-7b5032a11a3e?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/7b5032a11a3e</guid>
            <category><![CDATA[math]]></category>
            <category><![CDATA[algebra]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[linear-algebra]]></category>
            <category><![CDATA[matrix]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Mon, 28 Sep 2020 15:59:54 GMT</pubDate>
            <atom:updated>2020-09-28T20:53:02.181Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fJd-OVRk0qgY-Ip1AT5mwg.jpeg" /></figure><p>Let’s say that we have a missile travelling. The missile is using sensors to determine its position. The sensor data are stored in matrices. Every new, for example, nanosecond a new reading from the sensor arrives. It is really expensive to recompute the whole matrix of the readings for calibrations and other operations.</p><p>When the right-hand side of linear system changes but the matrix stays the same, then an LU factorization is needed to avoid computing the whole system again.</p><h4>But what if the matrix changes?</h4><p>It is possible to avoid refactoring even with a change in the matrix. For example, making a <strong><em>“Rank-one update”</em></strong>. That<strong><em> </em></strong>means adding to our matrix another matrix of rank one.</p><h4>How we will avoid computing the inverse of the whole matrix?</h4><p>The trick is using the Sherman-Morisson-Woodbury formula. The formula will give us the inverse of the matrix that results from a rank-one update but without computing the inverse multiple times.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nJX_5YTBQafW8u3t4mTtnQ.png" /><figcaption>The Sherman-morisson formula, where u and v are n-vectors.</figcaption></figure><p>It is clear that the inverse of A can be calculated only once. This allows us to work with matrix products that need O(n²) rather than computing the inverse in O(n³).</p><h4>Solving the Linear System</h4><p>Suppose we have a change in our matrix. We can obtain the following system:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/818/1*QocSd9Od5h8oVC0aNjJi-A.png" /></figure><p>And by using the Sherman-Morisson-Woodbury formula we obtain:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2dcXK9NYq2XEM94_tnfTuw.png" /></figure><p>We can then follow the steps bellow:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0Gp0u-KMsayy-iNOb-uNGg.png" /></figure><h4>Let’s look at an example</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zq4kp7rTzPYC77U7zvvsow.png" /></figure><p>And now a small change occurs in position (3,2). From -3 we have -1.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qTCrzW98ldvFHSaiwJRMrw.png" /></figure><p>We now need two update vectors so that the following stands.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_m-wupzm0JWRqhZt_jJNuQ.png" /></figure><p>These two work for our example.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Y_Lmrc8dhodF_52lhL9DKw.png" /></figure><p>Now lets compute step 1 and 2.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*o0Yc8M3m5oMGHBbiQo66PQ.png" /></figure><p>And finally, solve the linear system without the need to recompute the new matrix every time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pFxpsM_UeFDUUtj9h7FD3w.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7b5032a11a3e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Another one-liner package broke the JS ecosystem]]></title>
            <link>https://itnext.io/another-one-liner-package-broke-the-js-ecosystem-c5c7fe061e08?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/c5c7fe061e08</guid>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[javascript-development]]></category>
            <category><![CDATA[programming-languages]]></category>
            <category><![CDATA[npm]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Sun, 26 Apr 2020 12:15:42 GMT</pubDate>
            <atom:updated>2020-04-26T13:28:28.147Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gE01OjzFzA5eCmtOqpQ4nQ.jpeg" /><figcaption>Photograph by Maël Ball</figcaption></figure><p>People just hate javascript to fit in, to look cool or to sound like they know what they’re talking about. This issue is one amongst many arguments that make the “JS bad”.</p><p>Javascript depends on npm packages. As we can see in this <a href="https://github.com/then/is-promise/issues/13">issue</a> this utility function, whose sole purpose is to return a boolean value, according to naming conventions, does not always return a boolean value.</p><p>We can see at the <a href="https://github.com/then/is-promise/network/dependents">dependency graph</a> that nearly 3.4 million repositories were affected. This is the one-liner that did the job.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/fcb2670111f57eb00cd1d242afdb393f/href">https://medium.com/media/fcb2670111f57eb00cd1d242afdb393f/href</a></iframe><p>Pretty unreadable, amirite? If you write it out in long form it’s more like 3 if statements and two returns and a bunch of edge cases. In javascript this is a reason for creating an entire package.</p><p>The main argument against it is that all the environments need to support it. In addition a lot of these one liners address deficiencies of the weak and dynamic type system. So the package manager(NPM) can be seen sort of as a bug patcher.</p><p>Why they don’t create a standard library? things would be faster in the browser and more reliable. As i said before environment support but also this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*9nMBMt-OugnruBr_M-WuEQ.png" /><figcaption>xkcd standards</figcaption></figure><p>Another problem is that this one-liners are used to get big stats in the NPM’s user rankings by posting a lot of them. Lots of people tend to brag about how their 1-line package was downloaded thousands or millions of times on their resume. Things like this make me consider leaving software engineering and instead find purpose in life with an indoor small dog hotel for corgis.</p><p>Different people want to build different things, they may not like JS as it is. Thats why they built an ecosystem of one-liners that everyone can create on top of it the “standard” that they have in mind.</p><p>Summing it up we have a “3.6 Roentgen not great not terrible” kind of language. Let’s hope it doesn’t blow up.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c5c7fe061e08" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/another-one-liner-package-broke-the-js-ecosystem-c5c7fe061e08">Another one-liner package broke the JS ecosystem</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bulk loading R-trees and how to store higher dimension data]]></title>
            <link>https://medium.com/analytics-vidhya/bulk-loading-r-trees-and-how-to-store-higher-dimension-data-c7da26e4f853?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/c7da26e4f853</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[spatial]]></category>
            <category><![CDATA[relational-databases]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Fri, 24 Apr 2020 14:08:19 GMT</pubDate>
            <atom:updated>2020-05-25T15:56:14.580Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mxQP-nJMK4vja2lHkfkKTQ.jpeg" /><figcaption>Photograph by Laura Stanley</figcaption></figure><h3>R-trees? More like B-trees on steroids</h3><p>R-trees are tree data structures used for spatial access methods, i.e., for indexing multi-dimensional information such as geographical coordinates, rectangles or polygons.</p><h4>Spatial data</h4><p>An object is characterized as spatial if it has at least one attribute that captures its location in a 2D or 3D space. Moreover a spatial object is likely to have geometric extent in space. For example, we can say that a building is a spatial object, since it has a location and a geometric extent in a 2D or 3D map.</p><p>There is no total ordering in the multidimensional space that preserves spatial proximity. As a result, objects in space cannot be physically clustered to disk pages in a way that provides theoretically optimal performance bounds to spatial queries.</p><p>Assume that we try to sort a set of two-dimensional points using some one-dimensional sorting key. No matter which key we use, sorted by their column a Peano curve or a Hilbert ordering, we cannot guarantee that any pair of objects, which are close in space, will also be close in the total order.</p><p>R-trees are mapping our world and this is a real world application. They store spatial data objects such as shop locations and every other shape that creates a map. For example the geographical coordinates that create a polygon from a city block or your favourite restaurant.</p><h4>Why this is useful?</h4><blockquote>Hey Siri! find the closest McDonald’s from me</blockquote><p>That’s right. Queries according to the actual location of an object.</p><h3>The “R” in R-tree stands for rectangle</h3><p>The key idea of the data structure is to group nearby objects and represent them with a rectangle. The <strong><em>minimum bounding rectangle </em></strong>or<strong><em> MBR </em></strong>for short. This happens recursively.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/749/1*Rg8GGRXhkS6TcSYMko6emg.png" /></figure><p>Since all objects lie within this bounding rectangle, a query that does not intersect the bounding rectangle also cannot intersect any of the contained objects. At the leaf level, each rectangle describes a single spatial object. All the other levels are simply node pointers.</p><p>A constraint of the tree is that each node (except from the root) should be at least 40% full so that disk space is utilized properly. Below we can see an example of an R-tree.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_wa85gISlshaiRxgfQ47Lw.png" /></figure><p>There are four things that we have to do in order to keep our R-tree utilized.</p><ol><li>The area covered by the MBRs of the directory node entries should be minimized.</li><li>The overlap between MBRs of the directory entries at the same level should be minimized.</li><li>The margins of the MBRs of the directory node entries should be minimized.</li><li>The average number of entries in an R–tree node should be maximized.</li></ol><h3>Bulk loading</h3><p>Bulk loading is a way to load data, typically into a database, in ‘large chunks’. We will use a method that is called<strong><em> sort tile recursive</em></strong> <strong><em>(STR)</em></strong>.</p><p>Every MBR contains the following separated by a tab(‘\t’): object-id, x-low, x-high, y-low, y-high. Below we can se a function that computes the MBR of a list of rectangles. Each rectangle is stored in a string.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9ff22dff361e706a8020062e1f70cb1b/href">https://medium.com/media/9ff22dff361e706a8020062e1f70cb1b/href</a></iframe><p>The <strong><em>STR</em></strong> technique initially sorts the n rectangles with respect to the x-low of the rectangle. After sorting, we know that there will be <strong>L = ⌈n/C⌉</strong> leaf nodes in the tree. Attribute <strong>C</strong> refers to the capacity of each node. Specifically how many rectangles it can store.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/3ae773146e7e5211f4c15c7091114b24/href">https://medium.com/media/3ae773146e7e5211f4c15c7091114b24/href</a></iframe><p>The (sorted) rectangles are then divided into <strong>⌈SquareRoot( L)⌉</strong> groups (vertical stripes) and each group is sorted using as a key the y-low of the rectangle. The output from this sorting is packed into leaf nodes and the tree is built in a bottom-up way with recursion calls.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/a98680aab54d4812ab14add1f8fe0ba2/href">https://medium.com/media/a98680aab54d4812ab14add1f8fe0ba2/href</a></iframe><p>The recursion is over when the number of slices is 1. That means we reached the root level.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/ff66ee518bc0acc883e673463019b2cd/href">https://medium.com/media/ff66ee518bc0acc883e673463019b2cd/href</a></iframe><h3>Lets query the tree</h3><p>The tree is traversed in a depth- first manner, following pointers of entries which intersect the query range. The query algorithm takes three parameters; the query range q, the predicate θ the retrieved objects should satisfy with q, and an R–tree node n.</p><p>If we are on a leaf node we search every MBR of this node to check if they satisfy our query. If they do we add the rectangle to the answer set.</p><p>If we are not on a leaf node we we search every MBR of this node to check id they satisfy our query. If they do we recursively call the function. We use the same parameters. The only parameter that changes is the node. We are now passing the node that is pointed by our current one.</p><p>In simple words we want to check if we have an intersection with the MBR of the query.</p><p>We want to be able to do 3 query types. <strong><em>Intersection,</em></strong> for MBRs that have at least one common point, <strong><em>Inside,</em></strong> for MBRs that are inside the query and <strong><em>Containment,</em></strong> for MBRs that contain the MBR of the query.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/fb54a7ae4736c556e37dfaa134cbd01d/href">https://medium.com/media/fb54a7ae4736c556e37dfaa134cbd01d/href</a></iframe><p>Every non-leaf node must be checked only for intersection. We can see below with an example that we would miss out a lot of results if we strictly filtered all MBRs on non-leaf nodes.</p><p>Consider an inside query and a step at a non-leaf node. The black rectangle is the MBR of the query. The red is an intermediate node being filtered and the green of an actual object. If in the recursion we had a strict rules for the MBRs to be completely inside to be examined the red would not pass the filter so we would loose the green one that is an actual object that fit the criteria.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/490/1*16FjiaN74puJfzBoND9UNw.png" /></figure><p>The criteria for accepting an object depends on its geometry and can be seen below.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c72d312fd9b42ae1aa487f87acf05317/href">https://medium.com/media/c72d312fd9b42ae1aa487f87acf05317/href</a></iframe><h3>Conclusion</h3><p>The way spatial selections are processed in a spatial database systems depends on whether the queried relation R is indexed. If R is not indexed, we scan it linearly and compare each object with the query range. As discussed, the DBMS applies the two-step processing technique, which tests the object’s MBR against the query, before its exact geometry. If the relation is indexed (e.g., by an R–tree) then we can use it to find fast the objects that qualify the filter step.</p><p><strong>References:</strong><em> Advanced Database Technologies</em> <em>Lecture notes by Nikos Mamoulis, University of Hong Kong</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c7da26e4f853" width="1" height="1" alt=""><hr><p><a href="https://medium.com/analytics-vidhya/bulk-loading-r-trees-and-how-to-store-higher-dimension-data-c7da26e4f853">Bulk loading R-trees and how to store higher dimension data</a> was originally published in <a href="https://medium.com/analytics-vidhya">Analytics Vidhya</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Basic principles of networking (Part 1-Introduction)]]></title>
            <link>https://leobouts.medium.com/basic-principles-of-networking-96cf05af725?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/96cf05af725</guid>
            <category><![CDATA[introduction]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[computers]]></category>
            <category><![CDATA[networking]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Fri, 27 Mar 2020 14:04:59 GMT</pubDate>
            <atom:updated>2020-04-24T14:15:19.349Z</atom:updated>
            <content:encoded><![CDATA[<h3>Taking notes on Networking Part-1</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NXTtG1X0Jo6afXI5FsQvdQ.jpeg" /><figcaption>Photograph by Timo Volz</figcaption></figure><h3>What is a computer network?</h3><blockquote>A set of two or more computers that are connected between them with one or more physical means.</blockquote><p>Every computer is called a <strong><em>node</em></strong>,<strong><em> </em></strong>a node of the network can be any kind of computer or a terminal. Every node is defined by at least one alphanumerical value that is called an <strong><em>address.</em></strong></p><p>A physical mean is called a <strong><em>link</em></strong>, <strong><em>channel</em></strong> or <strong><em>couple</em></strong>. For example, optical fibers or coaxial cables.</p><p>The nodes and links make up the resources of the entire network.</p><h4>Point-to-point and Full Mesh</h4><p>The simplest way to connect is a direct link between two nodes. Now if we have <strong><em>N</em></strong> nodes and want to connect everyone with everyone we will need a number of <strong><em>N*(N-1)/2 </em></strong>connections. This network is called a<strong><em> full mesh </em></strong>network.</p><h4>Networking challenges</h4><p>What is <strong><em>scalability</em></strong>?</p><blockquote>The ability of a network that gets bigger but remains efficient and cost-effective.</blockquote><p>Full mesh networking is not appropriate for big scale networks. This is the result of financial, technological and physical constrains. Connecting <strong><em>N*(N-1)/2</em></strong> nodes is not an easy task. This number gets bigger as the network grows and physical links are expensive. The cost of the links is increasing non linearly in respect of its length. For example Km of optical fiber. In addition every computer would have to deal with a big load caused of all the links.</p><p>For a much more efficient network we need to constrain the number and length of the links and also use some of the nodes only for interfaces. Eventually we need to jointly share system resources(links and nodes) and have indirect connectivity with the help of partner nodes.</p><h3>Direct link networks</h3><p>Basic idea: connect while using only one shared link.</p><p>Every node will try to access the shared link. This node will be the only node that will use the link. When the node gets access it gets the ability to communicate with another node directly with a simple information transmission. This communication is called a <strong><em>one hop</em></strong>. These networks are called <strong><em>direct link</em></strong> networks, <strong><em>multiple access</em></strong> networks or <strong><em>transmission </em></strong>networks.</p><p>There are different kinds of direct link networks. In respect to the physical link (coaxial cable, fiber, wireless) or how the nodes are connected, the topology of the network(bus topology, ring topology, etc).</p><p>Direct link networks are simple to make but the number of nodes that can be connected is limited due to technological and physical constrains. The bigger the number of nodes the slower the network becomes. In addition we have signal reduction as distance get bigger and an increase in <strong><em>signal propagation delay</em></strong>.</p><blockquote><strong><em>Direct link networks do not cover all of our networking needs.</em></strong></blockquote><h4>The network switch</h4><p>The basic idea of switches lay in the shared system resources. The node that offers the networking services is called a <strong><em>switch </em></strong>and the node of a user is called a <strong><em>host. </em></strong>Hosts do not communicate directly with themselves. The switch receives the information from the sender and transfers it to the receiver.</p><p>A switch offers networking to a limited number of hosts. A lot of switches can be connected together to support a bigger number of hosts. This results to the <strong>switch network</strong>. The communication of the hosts is achieved with more than one hops(multihop communication). In other words switches transfer information from link to link.</p><p>In a switch network the switches and the links that connect them are shared resources and form the <strong>backbone</strong> network.</p><h4>The Internet</h4><p>Switch and Direct networks can be connected with switches to implement an <strong><em>internet</em></strong>. This is important because we need to connect heterogeneous networks together, reach a bigger number of hosts and connect in different geographical areas.</p><p>There is a direct connection of the technical side of a network in respect of the geographical area it covers.</p><p>We can categorise networks as <strong><em>Local Area Networks(LAN) </em></strong>if we refer to direct link networks that are no lengthier than 100m. <strong><em>Metropolitan Area Networks(MAN) </em></strong>that more commonly are switch networks spreading through entire cities. <strong><em>Wide Area Networks(WAN) </em></strong>that are switch networks and internets with world-wide coverage.</p><h4>What is a computer network again?</h4><blockquote>A set of two or more <strong><em>computers or networks</em></strong> that are connected between them with one or more physical means and <strong><em>routers/switches</em></strong>.</blockquote><h4>Network tradeoff</h4><p>Networking techniques allow us under certain conditions to implement big scale networks without a full mesh topology. In return we have to design and implement a process with increased complexity. For example we must have rules for sharing the network resources and the network should operate in a distributed way.</p><h4>Design of a network</h4><p>Networking services must be provided for a big number of computers while remaining cost-efficient. In addition a network must provide services like p2p file sharing and support new types of wireless networking.</p><p>We can reduce the design complexity by analysing all the individual mechanisms of a network. A.k.a. <strong><em>abstractions</em></strong>.</p><p>Every abstraction should be independent of the other, hiding the implementation details of its mechanism and determine the services the mechanism is offering.</p><p>In telecommunications the relations of individual mechanisms leads in the <strong><em>layering</em></strong> of abstractions. Every <strong><em>layer </em></strong>is called a<strong><em> level</em></strong>. The number of layers is defined from the architect of the network. Layering reduces the design complexity and gives flexibility in design changes. On the other hand the design might not be optimized for the efficiency of the network and in some cases it is impossible to achieve independence between the layers.</p><h4>And what about network protocols? what is a protocol?</h4><blockquote>All the rules that implement one or more mechanisms of a network level and provide a communication service.</blockquote><p>Every protocol defines a service interface. This interface contains all the communication services that the upper levels can use. Also protocols define a peer interface. This interface is about the structure of the messages that two <strong>same </strong>protocols exchange in different computers.</p><p>Every level can have more than one protocols(TCP, UDP) and the connections between the protocols can be represented with a protocol graph.</p><blockquote>The <strong><em>network architecture</em></strong> is defined as all the rules that configure the structure and content of a protocol graph.</blockquote><h4>Encapsulation</h4><p>In a switch network applications create data in sections that called <strong>messages</strong>. Every protocol inserts a <strong><em>header</em></strong> in the message. The header contains information on how this message will be handled by the same protocol in the other computer. In some cases this information is inserted in the end of the message and called a <strong><em>trailer</em></strong>. This process is called <strong>encapsulation</strong>. The section of the information that a protocol receives is called a <strong><em>payload</em></strong>.</p><blockquote>The payload contains information that the protocol cannot understand</blockquote><p>In the destination computer every level reads the header of the corresponding header and forwards the payload to the upper level protocol.</p><p>Generally network data is called</p><ol><li><strong><em>Message</em></strong> at the application level</li><li><strong><em>Packet </em></strong>at the network level</li><li><strong><em>Frame</em></strong> at the link level</li></ol><h4>Multiplexing and demultiplexing</h4><p>A protocol can provide services in more than one upper level protocols. For example the IP protocol forwards messages that are coming from TCP or UDP. A protocol must also forward every message that is coming from the upper level. This process is called <strong><em>multiplexing</em></strong>.</p><p>The header contains an id that is called the demultiplexing key. This id defines from which protocol the data came. In the receiver node the corresponding protocol uses the demultiplexing key to deliver the message to the right protocol. For example the IP uses this key to deliver the message to the UDP or the TCP protocol.</p><h4>The OSI model</h4><blockquote>Open Systems Interconnection</blockquote><p>The OSI model is defined by 7 levels. This is just a specification and <strong>not </strong>a network architecture.</p><p>From the lower to the highest level the OSI model contains these levels of abstraction:</p><ol><li>The <strong>Physical</strong> layer. Manages the transport of the information through a physical mean and describes the physical mean that is being used.</li><li>The <strong>Data Link</strong> layer. Defined the rules and how the shared resources will be used. Also implements the technical details for a reliable information transmission.</li><li>The <strong>Network</strong> layer. Describes the transmission and the routing of the information but also the mechanisms that routers use.</li></ol><p>The three lower levels are implemented in every node of the network, hosts and switches/routers. The other levels are only implemented in hosts.</p><p>4. The <strong>Transport</strong> layer. Describes the reliable transportation from process to process and provides an interface so that applications can connect with the network.</p><p>5. The <strong>Application</strong> layer. Contains protocols for implementing internet apps like smtp, ftp, http etc.</p><p>6. The <strong>Presentation</strong> layer. Describes how the data are formatted so that there is no problem with different operating systems.</p><p>7. The <strong>Session</strong> layer. Describes the rules for establishing a connection and synchronize between processes.</p><h4>The TCP/IP</h4><p>This is the technique that the Internet uses and is described with 4 levels.</p><p>The Application level that contains all the internet apps that are essential. FTP, SSH, SMTP etc.</p><p>The Transportation level that describes the communication between the network. The most important protocols here are the TCP(Transmission Control Protocol) and UDP(User Datagram Protocol).</p><p>The Internet level that describes the basic operations like routing. (IP)</p><p>And finally the connection of computer-network level that describes the technologies that connect computers to the network. For example Ethernet or wireless.</p><h4><strong>The Building Blocks</strong></h4><p>Information is transmitted through a channel as a <strong><em>digital signal</em></strong>. A digital signal is a signal that can only get discrete values. A lot of information sources on our computers create <strong><em>analog signal</em></strong>. For example recording your own voice. Analog signals can have continuous values. Analog signals can be transmitted through a channel but it is better to digitise them first. Why? <strong>Noise happens</strong>.</p><p>The analog signal can be easily altered through channel noise. Various reasons for that, for example the weather! Digital signal are resistant to noise because the get discrete values and can be reconstructed. Even with a digital signal the reconstruction is limited and depends on the levels of power and noise of the signal. Digital signals allow the information to travel in a greater distance with transmitters boosting its signal and isolating noise and gives us privacy by encrypting the signal.</p><p>The encryption techniques can differ in the number <strong><em>V</em></strong> of the discrete values (or symbols) a signal can have or in the <strong><em>duration</em></strong> of the values <strong><em>Ts </em></strong>that define the<strong><em> baud rate.</em></strong></p><blockquote>A signal with <strong><em>V</em></strong> symbols and a <strong><em>baud speed</em></strong> of <strong><em>1/Ts</em></strong> symbols per second can carry information with a rate of <strong><em>logV / Ts</em></strong> bits per second.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=96cf05af725" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A facility location problem. Where is the optimal placing?]]></title>
            <link>https://medium.com/swlh/a-facility-location-problem-where-is-the-optimal-placing-39f034ac1332?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/39f034ac1332</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[java]]></category>
            <category><![CDATA[dijkstra]]></category>
            <category><![CDATA[graph-theory]]></category>
            <category><![CDATA[computer-science]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Tue, 22 Oct 2019 18:13:13 GMT</pubDate>
            <atom:updated>2019-10-23T15:06:59.664Z</atom:updated>
            <content:encoded><![CDATA[<p>The study of facility location problems (FLP), also known as location analysis, is a branch of operations research and computational geometry concerned with the optimal placement of facilities to minimize transportation costs while considering factors like avoiding placing hazardous materials near housing, and competitors’ facilities.</p><p><strong><em>Consider the following scenario</em></strong></p><p>A rural area of a country has decided to improve their health and medicine access. They decided to build a new hospital and a pharmacy but everyone argues about the optimal location of these facilities. Some argue about the population of the cities and others about the distance.</p><p>The first step is to create a model of the rural area so we can apply some logic. Using graph theory we will create a non-directed graph of the cities of our rural area. Each vertex will be a city/village and the edges are the connections between them. For example let’s take the following graph.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/421/1*5UcTzpl82w4LnRoY6LJVhg.png" /><figcaption>Graph representing the cities of the rural area</figcaption></figure><p>Before we try to find the optimal location of the hospital/pharmacy lets define some graph theory terms.</p><h4>Eccentricity</h4><p>Eccentricity is the maximum distance from vertex 𝑣 to any other vertex<br>𝑢 of the graph. e(v)= max{𝑑( v, u) | 𝑢 ∈ 𝑉(𝐺)}. For example the Eccentricity of the node 3 is 3 if we consider that every edge has a weight of 1.</p><h4><strong>The Center of the Graph</strong></h4><p>The subgraph of a graph 𝐺 induced by the vertices of 𝐺 with the minimum<br>eccentricity is called Center of graph G (denoted as center(𝐺)).</p><p><strong><em>In our graph we can see clearly that the center of the graph is the node 3.</em></strong></p><h4>Vertex Distance</h4><p>Vertex Distance is defined as the sum of the distances between a specific<br>vertex and all the vertices of the graph.</p><p>distance of 𝑣 = SUM (𝑑𝑖𝑠𝑡(𝑣, 𝑢)) for every 𝑢 ∈ 𝑉(𝐺)</p><h4>The Median of the Graph</h4><p>The subgraph of a graph 𝐺 induced by the vertices of 𝐺 with the minimum<br>distance is called Median of graph 𝐺 (denoted as median(𝐺)).</p><p><strong><em>In our case the median of the graph is the node 6.</em></strong></p><h3>What about the hospital? what about the pharmacy? Where we should put them?</h3><p>The hospital must be near the quickest routes. <strong>Take note that an ambulance has to make the same path twice</strong>, one dispatching the vehicle to the location and one returning the patient back to the hospital.</p><p>In the other hand the pharmacy just has to be accessible by the citizens.</p><blockquote>The median of the graph can minimize the time because it has the minimum eccentricity. Thats where the hospital must be placed. As you imagined, the center of the graph can be used for the pharmacy because it’s more accessible by everyone.</blockquote><h3>Let’s code it.</h3><p>First thing to do is to transfer the graph in some beep boop computer language. We will use Java.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c98b7e0c849bc26ff69868d944409b88/href">https://medium.com/media/c98b7e0c849bc26ff69868d944409b88/href</a></iframe><p>We need to use the Dijkstra algorithm so we can find the all the shortests paths for every node.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c7c830b5bd72c5e25c96d88929245769/href">https://medium.com/media/c7c830b5bd72c5e25c96d88929245769/href</a></iframe><p>These two helper functions will be used by the Dijkstra algorithm to find the minimum and maximum value of the array.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/89ecaece13ea1fadb2d159da239976a9/href">https://medium.com/media/89ecaece13ea1fadb2d159da239976a9/href</a></iframe><p>After getting the paths we can find the eccentricity the center and the median as follows</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e5de3c58f1e17a3168d8740ffe69b404/href">https://medium.com/media/e5de3c58f1e17a3168d8740ffe69b404/href</a></iframe><p>The main of the class will be as follows.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/15a4b36ff54235328f39b5207dc5d5f9/href">https://medium.com/media/15a4b36ff54235328f39b5207dc5d5f9/href</a></iframe><h3>Conclusion</h3><p>We checked a simplified version of a problem that can be very tedious if many factors are acknowledged and connected it with graph theory. At first we made a model of the rural cities with a graph. Using the center and the median of a graph we created an automated way of placing these two facilities in the optimal location. Graph theory helps us model and solve problems using existing theorems and applying everyday logic to them. Every application of the graph theory properties depends solely on the problem we’re facing, in our case the hospital and the pharmacy locations.</p><p>You can check the live demo of the code <a href="https://repl.it/@nikosp2196/GraphsTheoryLab1">here</a></p><p>This blog post was made to feed and share our curiosity in collaboration with <a href="https://medium.com/u/be8e6a6c9f28">Nikos Pantelidis</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=39f034ac1332" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/a-facility-location-problem-where-is-the-optimal-placing-39f034ac1332">A facility location problem. Where is the optimal placing?</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Build a search engine: Searching the Yelp dataset using SOLR , Angular and Node]]></title>
            <link>https://itnext.io/build-a-search-engine-searching-the-yelp-dataset-using-solr-angular-and-node-1cd256e7ac24?source=rss-9bae4e527b64------2</link>
            <guid isPermaLink="false">https://medium.com/p/1cd256e7ac24</guid>
            <category><![CDATA[solr]]></category>
            <category><![CDATA[node]]></category>
            <category><![CDATA[angular]]></category>
            <category><![CDATA[api]]></category>
            <category><![CDATA[search-engines]]></category>
            <dc:creator><![CDATA[Leonidas Boutsikaris]]></dc:creator>
            <pubDate>Fri, 26 Jul 2019 15:38:44 GMT</pubDate>
            <atom:updated>2019-08-01T12:53:43.928Z</atom:updated>
            <content:encoded><![CDATA[<h3>Build a search engine: Searching the Yelp dataset using SOLR , Angular and Node</h3><p>In this post we’ll get through the process of creating a search engine app with Angular. We will check out how solr works and the simplest way to configure our index based on our dataset.</p><blockquote>You can browse the code of the project while reading through this blog <a href="https://github.com/leobouts/yelp.ene/tree/master/src">HERE</a>.</blockquote><p>Let’s start by defining all the slang we need to know.</p><h3>What is an index?</h3><p>An index is a list of data. It is typically saved in plain text so that can be quickly accessed by a search algorithm. This significantly speeds up searching. Indexes often include information about each item in the list, such as metadata or keywords, this allows the data to be searched via the index instead of reading through each file individually.</p><h3>And what about SOLR?</h3><p>Solr is an open-source search platform, written in Java. Solr runs as a standalone full-text search server. It uses the Lucene Java search library at its core for full-text indexing and search, and has REST-like HTTP/XML and JSON APIs that make it usable from most popular programming languages.</p><h4><em>…yeah that’s nice but where are the data?</em></h4><p>As mentioned before we’re gonna use the yelp dataset that can be found <a href="https://www.yelp.com/dataset"><em>here</em></a><em>.</em></p><p>The dataset contains many files that we can use but we are gonna need only three of them.</p><p>“business.json, review.json and tip.json”</p><p>Below we can see all the fields they contain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cFRkJtKSY8XZXpoe4581Mw.png" /><figcaption>fields of the json files</figcaption></figure><h3>Preparing the data</h3><p>We will use only a portion of the dataset. The script bellow will choose 100 random categories and select businesses until we reach 10000 of them.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e9d549ec86506a151346a8ce348332e6/href">https://medium.com/media/e9d549ec86506a151346a8ce348332e6/href</a></iframe><p>Now let’s continue by finding out all the reviews users left for the 10000 businesses we chose.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/dd056a557cc0c0ccb930053d66019b1d/href">https://medium.com/media/dd056a557cc0c0ccb930053d66019b1d/href</a></iframe><p>Finally we will merge businesses, reviews and tips in a final.json file so we can index it without much trouble.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0c665cf6f6dc7323c81209f0620f29e1/href">https://medium.com/media/0c665cf6f6dc7323c81209f0620f29e1/href</a></iframe><blockquote>The fields of the final.json file can be seen below.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ORivR1kIXUn9az7muKTdAA.png" /><figcaption>final.json fields</figcaption></figure><p>The reviews field is a json list with every json containing this info: stars, text, date. Also the attributes field references the tips file that we mentioned in the beginning.</p><p><strong><em>In the dataset i found an inaccuracy on the review_count field. The review_count field of business.json is not the exact number of reviews for some of the businesses(reviews.json file). There is a function in the business_filter.py that proves this.</em></strong></p><h3>Setting up SOLR</h3><p>You can easily get apache solr from <a href="https://lucene.apache.org/solr/downloads.html">here</a></p><blockquote>Fun fact: SOLR stands for “Searching On Lucene with Replication”.</blockquote><p>Solr creates the so-called search cores that are separate index pieces. For example, if we have the core named core1, a typical search would be</p><blockquote>http://localhost:8983/solr/core1/select?Q =’query’</blockquote><p><strong>To run queries or to create a new core the solr server has to be fired up.</strong></p><blockquote>./bin/solr start</blockquote><p>For our app we will use a core name demo. Cd into the solr files and create it like this.</p><blockquote>./bin/solr create -c demo</blockquote><h4>Configuring SOLR</h4><p>Solr uses 2 configuration files. The managed-schema and solrconfig.xml. The solrconfig.xml file includes settings for the data directory location, cache parameters, request handlers etc. The managed-schema determines what kind of fields will be included in the search, which of these fields will be used as a single / primary key, which fields are mandatory and also how to search and index in each field.</p><p><strong><em>A primary key is a unique identifier for a record.</em></strong></p><p>For the solrconfig.xml file we will use a minimal version of the config.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V2-EddKNznryuz4dt-jl4g.png" /><figcaption>solrconfig.xml</figcaption></figure><p>For the managed-schema file, we will use the default configuration sets given by the solr. We will also erase several lines of code used for languages other than English and only add the appropriate indexing fields we want to search.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pddPQV_7UmVFazaHEyJGmA.png" /></figure><blockquote><em>reminder: all these files exist in the github repo provided in the beginning</em></blockquote><p>We can see very easily that the fields in the managed-schema are the same with the fields in the final.json file that we are going to use. The business_id field is used as a unique key. The rest of them are going to define how the indexing will be made.</p><p>“names, categories, reviews, stars, review_count, address, city, state, and attributes”</p><h4>Loading data to SOLR</h4><p>Once we have finished with the solr settings, we can load data into our core with the command</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/a6fb32adfde011afb3d3995bec5ce404/href">https://medium.com/media/a6fb32adfde011afb3d3995bec5ce404/href</a></iframe><h4>SOLR queries</h4><p>We are now ready to ask questions on the Solr API. Questions are made based on the port that solr runs, the name of the core, and the fields of indexing. We can see a search example based on the name field and reviews with the keyword ‘chinese’.</p><blockquote><a href="http://localhost:8983/solr/demo/select?q=name%3Achinese%20AND%20reviews%3Achinese">http://localhost:8983/solr/demo/select?q=name%3Achinese%20AND%20reviews%3Achinese</a></blockquote><p><strong>where %3Α is the browser code for “ : ” and %20 for the space character.</strong></p><h3>..yeah that’s also nice but how that solr thing works?</h3><p>When a user searches on Solr, the query is processed by a request handler. The request handler is a plugin that defines the logic to be used when Solr processes a request. Solr supports a variety of request handlers. Some are designed to process search queries while others handle tasks such as index replication.</p><p>Search applications select the Solr’s default request handler. To process a query, the request handler calls an analyser, the analyser interprets the terms and parameters of a query. Solr’s default query analyser is known as the Standard Query Parser or known as the “lucene” standard analyser. Solr also includes the DisMax query analyser and Extended DisMax (eDisMax) query analyser.</p><p>Standard Query Parser syntax allows high accuracy searches, but the DisMax query analyser is much more tolerant of errors. DisMax is designed to provide a similar experience with popular search engines like Google, that rarely show syntax errors to users.</p><h3>Solr needs to talk to the app, the Node API</h3><p>In order to establish a communication between solr and the application we need an intermediate server that will communicate with both. It is especially necessary for the Cross-Origin Resource Sharing (CORS) functionality in browsers. We will use the express and cors libraries. The API will consists of 3 endpoints that generate questions based on what the user has chosen in the search engine.</p><p>you have to install <a href="https://nodejs.org/en/download/">node</a> and also</p><blockquote>npm install express</blockquote><blockquote>npm install cors</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TKflNDdIGdnYtenGsbTYlg.png" /></figure><p>The first endpoint serves basic types of questions. The second and third serve boolean questions. The first endpoint code makes an http call with the baseUrl parameter which is the local Solr address by matching the parameters given to it by the application, ie the keyword and the type of search. The other two endpoints are implemented accordingly.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d7188a5a4d504eeaef409819df31dbf5/href">https://medium.com/media/d7188a5a4d504eeaef409819df31dbf5/href</a></iframe><h3>Lets get to the App</h3><p>The application will be built using Angular and Angular Material available <a href="https://material.angular.io/guide/getting-started?source=post_page---------------------------">here</a>.<br>(Don’t forget any step in the angular material installation guide).</p><p>The first screen looks something like this.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/981/1*04nb2-rO-0WcdvJPWTlXxQ.png" /></figure><p>All the HTML code lies inside the app.component.html file. Two divs contain the whole view. One is the search view and the other is the search results view. with the *ngIf=”searchPageViewActive” we can navigate between the two divs.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/61f76da243f68986cae568f99bd8c4b6/href">https://medium.com/media/61f76da243f68986cae568f99bd8c4b6/href</a></iframe><p>Now the searchPageViewActive boolean variable will become false as a progress bar indicates that a query is being made and the result page will pop up after we have the data back.</p><p>The result page will look like this.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c1_yhBFJqRA1zYaYxFzo1g.png" /></figure><p>The results are ordered exactly as solr returned them. We can change that order according to the stars a business have. If there is a tie the review count will be considered as a second ordering parameter.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/bb53cac0c1f8373564493abc5a951bc8/href">https://medium.com/media/bb53cac0c1f8373564493abc5a951bc8/href</a></iframe><p>There is also a dialogue box that opens when we click the number of reviews that displays the full info of every review.</p><p>The code of the dialogue can be seen here.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/32bc961a5e0495ca01af9315f8028104/href">https://medium.com/media/32bc961a5e0495ca01af9315f8028104/href</a></iframe><h3>Before checking our functions lets get to the Data Service</h3><p>The DataService.ts file serves as an intermediate between our app and our Node API. Remember the three different query types from the api? We will using three different calls from our application to the API whenever is needed.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/fbbf884370e677c89a70ce9da4615d3a/href">https://medium.com/media/fbbf884370e677c89a70ce9da4615d3a/href</a></iframe><h3>Now the search functions</h3><p>Clicking the search button will invoke the search function that will validate the search keywords and syntax. If the syntax doesn’t make it, some snackbars will pop up displaying a warning. If the query is good to go, the search function will determine which of the three types the user chose to search, format the query to be solr ready with the cleanQuestion function and finally call the API through the DataService. Finally the cleanData function will prepare the data for display proposes to the Angular material table.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/15ab18d7dde72ac7e48f71caec0ba2ca/href">https://medium.com/media/15ab18d7dde72ac7e48f71caec0ba2ca/href</a></iframe><p>Bellow we can see how the dialogue handles and cleans the review data.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/cd710522ce0bf82f45787a8da0683c2c/href">https://medium.com/media/cd710522ce0bf82f45787a8da0683c2c/href</a></iframe><h3>Some Extra stuff..</h3><p>let’s do some <strong>word highlighting</strong>. We will create a pipe that takes a list of words as an input and highlights them in every place they appear on the html view.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/95ee3a27007fdc135f45739cbf93d193/href">https://medium.com/media/95ee3a27007fdc135f45739cbf93d193/href</a></iframe><h4>Why not add a second way of ordering the results?</h4><p>let’s keep track every click users made so we know which business caught their eyes. This will be done every time a review dialogue opens. Google <a href="https://firebase.google.com/">firestore</a> will be used for storing the data.</p><p><em>Firestore has detailed documentation and a lot of tutorials are available online, you should check them out.</em></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d8959477681fbddab5dc903d8104ecf8/href">https://medium.com/media/d8959477681fbddab5dc903d8104ecf8/href</a></iframe><h3>Summary</h3><p>At first we prepared the data with some python scripts. We then proceeded to set up and configure. Finally we created an API so our front-end app can talk to the server, prepare the data and display it to the search results view.<br>Building a complete application from scratch can be quite tedious , this forces you to learn a lot in the process. Even with a simple app you have to work with many tools before ending up with the desired outcome.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1cd256e7ac24" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/build-a-search-engine-searching-the-yelp-dataset-using-solr-angular-and-node-1cd256e7ac24">Build a search engine: Searching the Yelp dataset using SOLR , Angular and Node</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>