gesund.ai

gesundai

THE PROBLEM OF SEXUAL ABSTINENCE

From very early times it seems possible to trace two streams of opinion regarding women

THE BEGINNINGS OF THE MIND AND LANGUAGE

Pranks were a big deal on the first team I worked with at Google, but it was hard to think up harmless, doable stunts. Now that I’ve left Google and am working on my own, I decided to compensate by brainstorming some prank-worthy emails that you can send to your co-workers this April Fool’s Day. To spice up your water-cooler conversations this week, pick out an email below to send out to your office-mates. Find one that is just believable enough that your colleagues won’t realize they’re being pranked, until they do :) Happy hijinks! No more belt : No leather in the office, starting tomorrow Subject : HR just notified me that we have a new employee who starts tomorrow that is allergic to leather. To protect his health and productivity, all from the office. This includes bags, shoes, pants, belts, and any other leather accessories. Please leave your leather at home! Thanks for your support. Email leather will be prohibited Ho ho ho : Company-wide Christmas apparel tomorrow Subject : I just got notice from our CMO that we’ll be filming for the company Holiday greeting video tomorrow. All employees are expected to appear in the video, so please wear red and green and any other holiday apparel you own (Santa suits, Christmas sweaters, reindeer noses, and onesies encouraged). Email Gesundheit : New office policy Subject : This has been a particularly bad flu season, and we’re concerned about how disease affects productivity. After much discussion and review of recent research, leadership has decided to issue an office . If you need to sneeze, please go to the bathroom to avoid spreading germs and wash your hands after the sneeze. Do not sneeze at your desk! This new policy starts tomorrow. Email policy against sneezing Color blind : Help me keep my eyes well! Subject : After several years of getting increasingly bad eyes and stronger prescriptions, I’ve finally gotten a diagnosis. My optometrist informed me that I have a condition in which the color black is damaging my eyesight, sort of like an allergy. I’ll be wearing shaded glasses for the next couple of weeks while I undergo treatment. Please on any team documents or presentations (bright colors are okay). Email refrain from using black Social Good : Support my nephew! Subject : Hey friends — My nephew, Adrian, is running the “Fun Run” 5K for charity at his middle school this weekend. He chose to raise money for the (AACF) and is looking for sponsorships. Please consider donating $10-$40 for this important cause! Thanks for contributing. Email Americans Against Constipation Fund Save the planet : Pro-environmental change to our office supplies Subject : Hi everyone! Starting this week, our supply closet will no longer stock staples, paper clips, or tape, for environmental reasons. For fastening paper, we will provide organic glue. Thanks for helping us reduce our office footprint! Email Hardware : IT swap Subject : Due to recent security revelations, our office . Tomorrow, you’ll find that your desktop computers have been replaced with the Samsung Chromebook Plus. If you have a company laptop, please exchange it tomorrow morning before getting to work. Thanks to our IT and operations team for organizing the swap. Email switching to Chromebooks Say no to drugs : Improve mental health by switching to tea Subject : In light of research about the damaging effects of caffeine on mental health, we’ve decided to switch our coffee to herbal tea. Starting tomorrow, our office manager will brew herbal and black teas, available for free in the micro-kitchen. I encourage you to choose the decaffeinated option :) Coffee will no longer be offered. Email Trump card : New board member Subject : Starting tomorrow, we’ve have a new board member and advisor who will be touring the office. Mr. Rex Tillerson is an accomplished businessman and public servant who has recently decided to invest in our upcoming round. As always, please be courteous and helpful when he’s walking around! Email No shoes, no shirt Company swag for summer! Subject: In celebration of spring, we’re giving out company-branded swimsuits! Come to my desk tomorrow to pick up logo-ed trunks and bikinis. Email: Islander : Accepting transfer apps for new company office Subject : It’s important to our COO to support developing economies and facilitate open borders, so we have decided to ! Construction of the office will begin next week. The new South Pacific facilities will include beachfront conference rooms, high-end video conferencing technology, an on-staff masseuse, and local residences for employees. Please notify me if you’d like to be added to the transfer list. Email open an office in Bali Oktoberfest : Mandatory unconscious bias training Subject : A recent analysis showed that 60% of our employees have birthdays in October. This fraction is shockingly high, and a follow-up committee is investigating what’s going on. In the meantime, we’re asking everyone to complete to ensure that candidates born in other months of the year are treated fairly. Will be sending out the link shortly; please review by the end of the week! Email mandatory unconscious bias training Noise : Warning: Yodeling convention Subject : FYI, our office manager informed me this morning that there will be a happening tomorrow on the three floors below us. Please bring earplugs or headphones if the noise will bother you. Email yodeling convention Quiet Game : Hallway to be nap area Subject : To attract top tech talent and stay competitive in recruiting, company leadership has decided to convert the hallway to a nap area as a perk for employees. You’re welcome to enjoy 20 min-1 hour power naps in the nap area any time of day (pillows provided). Respect your peers by remaining silent in the hallway — no phone calls, conversations, or meetings, please. Email Buzzness : Space reorganization Subject : We’ve decided to reorganize our cubicles to achieve better organizational synergy. According to the recent findings of our McKinsey consultant, reshaping our work area into a will improve communication and collaboration across teams. We ask that you clear off your desk tomorrow morning to make it easier to move. Email pentagonal hive Twerk it : Fwd: Video of CEO twerking at Christmas party Subject : LMFAO I just found this in my saved Snaps Email Please hold : Plumbing outage tomorrow Subject : Plumbing in our building will be turned off from noon to 5. If you need a bathroom during that time, please use the toilet at the Subway down the street (I’ve already called them to make sure this is alright). Thanks and sorry for the inconvenience Email Take a hike : Please make alternative parking arrangements Subject : According to a new County code, our parking lot has been restricted as it is the breeding ground for a rare type of bird. Employees will no longer be able to park at our office building. Please plan for alternative means of transportation or a different parking lot. I’m sorry for the inconvenience. Email Nǐ hǎo : CTO request Subject : Our CTO has decided to start learning Mandarin so that he can better communicate with the engineering team. If you speak Mandarin, he’s asked that you talk to him exclusively in Chinese. He may act confused or ask you to repeat in English, but he has told us ahead of time to not go easy on him; he wants a truly immersive experience. Email IRL : Camera crews starting tomorrow Subject : I’m thrilled to announce that our office was chosen to be part of a reality show! Starting tomorrow, you’ll notice camera crews setting up and wandering around the office. The name/premise of the reality show is secret, but you don’t need to do anything special. Just behave normally and continue business as usual! Email Extra : Scent vendors Subject : Our office managers have decided to experiment with new odors for the office. Tomorrow, a vendor will bring samples of new scents and perfumes that we’re considering. Please stop by my desk to test out and vote on your favorites. Email Office antics in Silicon Valley Thanks all for reading! If you enjoyed, please share some Medium love and check out for , our online video editing website, for more articles. the blog Kapwing

Prank your co-workers with one of these emails

In the we dealt with some algorithmic issues. We were only able to solve these by improving the data structures in our code. First we used Haskell’s built in array type for fast indexing. Then when we needed a segment tree, and we decided to make it from scratch. But we can’t roll our own data structure for every problem we encounter. So it’s good to have some systems we can use for more of these advanced topics. last two articles One of the most important categories of data structures we’ll need for algorithms is graphs. Graphs are quite powerful when it comes to representing complicated problems. They are very useful for expressing relationships between data points. In this article, we’ll see two types of graph problems. We’ll learn all about a library called the (FGL) that is available on Hackage for us to use. We’ll then take a stab at constructing graphs using the library. Finally, we’ll see how simple it is to solve these algorithms using this library once we’ve made our graphs. Functional Graph Library For a complete set of the code that we’ll use in this article, check out . It’ll show you how you can use Stack to bring the Functional graph library into your code. this Github repository If you’ve never used before, it’s an indispensible tool for creating programs in Haskell. You should try out our and learn more about it. Stack Stack mini-course Graphs 101 For those of you who aren’t familiar with graphs and graph algorithms, I’ll explain a few of the basics here. If you’re very familiar with these already, you can skip this section. A graph contains a series of objects and encodes various relationships between these objects. For each object in our set, we have a “node” in our graph. These are like data points. Then to represent every relationship, we create an “edge” between two different nodes. We often give some kind of value to this edge as a piece of information about the relationship. In this article, we’ll imagine our nodes as places on a map, with the edges representing legal routes between locations. The label of each edge is the distance. Edges can be both “directed” and “undirected”. Directed edges describe a 1-way relationship between the nodes. Undirected edges describe a 2-way relationship. Here’s an example of a graph with directed edges: Making Graphs with FGL So whenever we’re making a graph with FGL, we’ll use a two-step process. First, we’ll create the “nodes” of our graph. Then we’ll encode the edges. Let’s suppose we’re reading an input stream. The stream will first give us the number of nodes and then edges in our graph. Then we’ll read a bunch of 3-tuples line-by-line. The first two numbers will refer to the “from” node and the “to” node. The third number will represent the distance. Here’s what this stream might look like for the graph pictured above: 691 2 31 3 42 3 52 4 22 5 63 5 54 6 95 4 15 6 10 We’ll read this input stream like so: import Control.Monad (replicateM)import System.IO (Handle, hGetLine)

data EdgeSpec = EdgeSpec  { fromNode :: Int  , toNode :: Int  , distance :: Int  }

readInputs :: Handle -> IO (Int, [EdgeSpec])readInputs handle = do  numNodes <- read <$> hGetLine handle  numEdges <- (read <$> hGetLine handle)  edges <- replicateM numEdges (readEdge handle)  return (numNodes, edges)

readEdge :: Handle -> IO EdgeSpecreadEdge handle = do  input <- hGetLine handle  let [f_s, t_s, d_s] = words input  return $ EdgeSpec (read f_s) (read t_s) (read d_s) Our goal will be to encode this graph in the format of FGL. In this library, every node has an integer identifier. Nodes can either be “labeled” or “unlabeled”. This label, if it exists, is separate from the integer identifier. The function we’ll use requires our nodes to have labels, but we won’t need this extra information. So we’ll use a newtype to wrap the same integer identifier. Once we know the number of nodes, it’s actually quite easy to create them all. We’ll make labels from every number from 1 up through the length. Then we represent each node by the tuple of its index and label. Let’s start a function for creating our graph: import Data.Graph.Inductive.Graph (mkGraph)import Data.Graph.Inductive.PatriciaTree (Gr)

…

newtype NodeLabel = NodeLabel Int type Distance = Int

genGraph :: (Int, [EdgeSpec]) -> Gr NodeLabel DistancegenGraph (numNodes, edgeSpecs) = mkGraph nodes edges  where    nodes = (\i -> (i, NodeLabel i))       <$> [1..numNodes]    edges = ... The graph we’re making uses a “Patricia Tree” encoding under the hood. We won’t go into details about that. We’ll just call a simple function exposed by the library. We'll make our return value the graph type parameterized by our node label type and our edge label type. As we can see, we’ll use a type synonym for integers to label our edges. mkGraph Gr Distance For now let’s get to the business of creating our edges. The format we specified with works out that we don’t have to do much work. Just as the labeled node type is a synonym for a tuple, the labeled edge is a 3-tuple. It contains the indices of the “from” node, the “to” node, and then the distance label. In this case we’ll use directed edges. We do this for every edge spec, and then we’re done! EdgeSpec genGraph :: (Int, [EdgeSpec]) -> Gr NodeLabel DistancegenGraph (numNodes, edgeSpecs) = mkGraph nodes edges  where    nodes = (\i -> (i, NodeLabel i))       <$> [1..numNodes]    edges = (\es -> (fromNode es, toNode es, distance es))      <$> edgeSpecs Using Graph Algorithms Now suppose we want to solve a particular graph problem. First we’ll tackle shortest path. If we remember from above, the shortest path from node 1 to node 6 on our graph actually just goes along the top, from 1 to 2 to 4 to 6. How can we solve this from Haskell? We’ll first we’ll use our functions from above to read in the graph. Then we’ll imagine reading in two more numbers for the start and end nodes. solveSP :: Handle -> IO ()solveSP handle = do  inputs <- readInputs handle  start <- read <$> hGetLine handle  end <- read <$> hGetLine handle  let gr = genGraph inputs Now with FGL we can simply make a couple library calls and we’ll get our results! We’ll use the module, which exposes functions to find the shortest path and its length: Query.SP import Data.Graph.Inductive.Query.SP (sp, spLength)

solveSP :: Handle -> IO ()solveSP handle = do  inputs <- readInputs handle  start <- read <$> hGetLine handle  end <- read <$> hGetLine handle  let gr = genGraph inputs  print $ sp start end gr  print $ spLength start end gr We’ll get our output, which contains a representation of the path as well as the distance. Imagine “input.txt” contains our sample input above, except with two more lines for the start and end nodes “1” and “6”: >> find-shortest-path < input.txt[1,2,4,6]14 We could change our file to instead go from 3 to 6, and then we’d get: >> find-shortest-path < input2.txt[3,5,4,6]15 Cool! Minimum Spanning Tree Now let’s imagine a different problem. Suppose our nodes are internet hubs. We only want to make sure they’re all connected to each other somehow. We’re going to pick a subset of the edges that will create a “spanning tree”, connecting all our nodes. Of course, we want to do this in the cheapest way, by using the smallest total “distance” from the edges. This will be our “Minimum Spanning Tree”. First, let’s remove the directions on all the arrows. We can visualize this solution by looking at this picture, and we’ll see that we can connect our nodes at a total cost of 19. The great news is that it’s not much more work to code this! First, we’ll adjust our graph construction a bit. To have an “undirected” graph in this scenario, we can make our arrows bi-directional like so: genUndirectedGraph :: (Int, [EdgeSpec]) -> Gr NodeLabel DistancegenUndirectedGraph (numNodes, edgeSpecs) = mkGraph nodes edges  where    nodes = (\i -> (i, NodeLabel i))       <$> [1..numNodes]    edges = concatMap (\es ->       [(fromNode es, toNode es, distance es), (toNode es, fromNode es, distance es)])      edgeSpecs Besides this, all we have to do now is use the function from the module! Then we'll get our result! msTree MST import Data.Graph.Inductive.Query.MST (msTree)

...

solveMST :: Handle -> IO ()solveMST handle = do  inputs <- readInputs handle  let gr = genUndirectedGraph inputs  print $ msTree gr

{- GHC Output>> find-mst < “input1.txt”[[(1,0)],[(2,3),(1,0)],[(4,2),(2,3),(1,0)],[(5,1),(4,2),(2,3),(1,0)],[(3,4),(1,0)],[(6,9),(4,2),(2,3),(1,0)]]-} This output is a little difficult to interpret, but it’s identical to the tree we saw above. Our output is a list of lists. Each sub-list contains a path from a node to our first node. A path list has a series of tuples, with each tuple corresponding to an edge. The first element of the tuple is the starting node, and the second is the distance. So the first element, refers to how node 1 connects to itself, by a single “edge” of distance 0 starting at 1. Then if we look at the last entry, we see node 6 connects to node 1 via a path through nodes 4 and 2, with total distance 9, 2, and 3. [(1,0)] Conclusion Graph problems are ubiquitous in programming, but it can be a little tricky to get them right. It can be a definite pain to write your own graph data structure from scratch. It can be even harder to write out a full algorithm, even a well known one like Dijkstra’s Algorithm. In Haskell, you can use the functional graph library to streamline this process. It has a built in format for representing the graph itself. It can be a little tedious to build up this structure. But once you have it, it’s actually very easy to solve many different common problems. Next week, we’ll do a bit more work with FGL. We’ll explore a problem that isn’t quite as cut-and-dried as the ones we looked at here. First we’ll take a more abstract problem and determine what graph we want to make for it. Then we’ll solve that graph problem using FGL. So check back next week on the !. Monday Morning Haskell blog The easiest way to bring FGL into your Haskell code is to use the tool. If you’re unfamiliar with this, you should take our free . You’ll learn the important step of how to bring dependencies into your code. You’ll also see the different components in a Stack program and the commands you can run to manipulate them. Stack Stack mini-course If you’ve never programmed in Haskell before, you should try it! Download our . It’ll point you towards some valuable resources in your early Haskell education. Getting Started Checklist

Stories, News & Mentions about GESUNDAI

Gesundai’s Stories on HackerNoon