You have probably all seen, at the Malta Phone Number List of confinement, this Forbes article ” What Do Countries With The Best Coronavirus Responses Have In Common?” Women Leaders ”saying in essence that countries led by women had fewer cases of Covid-19 or deaths from Covid-19 than others. The subject of parity between women and men in itself deserves attention, arouses reactions and only progresses through education. It starts with a press article, which is based on a study and was relayed on social networks. Everyone’s reaction can be “Finally, I agree, I retweet” or on the contrary “more nonsense”. Social networks are ruthless

on releasing information without recoil. So how can we take a step back from these important analyzes and announcements? Post of Forbes article on social media Photo from Forbes article (#girlspower) Having always been convinced that diversity is an important factor in team performance, I hastened to relay this post … Emoticon Not easy to resist social networks … After this impulsive click, the team (which I am training on Lean 6 sigma) kindly reminded me that I had been a victim of social networks and that this kind of statement had to be verified! Here I am again a student, using a basic statistical background and the big

That There Is A Linear Regression Which

data analysis tools that we provide to all our consultants. The first results that a consultant (Pierre) communicated to me amused me, and unfortunately not having been able to follow our own big data training modules, I had to learn on the job to redo his analyzes. And in doing so, it allowed me to flush out some mistakes made by Pierre, like what… it helps the experience! Data recovery and cleaning First: the source of the data. Here Pierre’s choice fell on the site www.worldometers.info I learned that it is very easy to retrieve data from a web page. For this, Google Sheets has a function called importhtml which allows you to

Malta-Phone-Number-List (1)

scrape any table (in particular) present on a web page and retrieve the data there. Very practical therefore, since the Covid-19 data evolves from day to day, this allows the subsequent analyzes made on this data to be quickly updated. importhtml allows to scrape a web page to retrieve data tables And there, while exploring the analyzes made by the consultant, in particular the aberrant points, I came across the data concerning Portugal which seemed to me… bizarre precisely. While going up the ball I ended up finding a problem (a bug of the importhtml function?). Indeed you will notice that on the worldometers site the data

It Is Simple To Calculate Some Additional

are in Anglo-Saxon representation with commas to separate the thousands (in France we use the space character as a separator of thousands). And that makes the importhtml function buggy. Look in the previous image for Chile (row 30) in the number of cases per million inhabitants (column I) the function returned the number 1.51 (instead of 1.510 so 1,510 in French notation) c ‘ ie 1,510 cases per million inhabitants. In short, all errors of this type must be corrected to correctly reconstitute the values ​​with large blows of formulas on the strings of characters. Strangely, I did not find people on the internet who had been

confronted with the same problem and posted a solution. First apprenticeship for me who had never done big data training: data cleaning is really complex and tricky (fortunately: when we have done data migration on large IS projects, we keep a few reflexes …) . Learning for Pierre, a consultant yet trained in big data: “verify, verify, and have the cleaning of your data verified by one of your peers or by an old schnock”. Pierre then wanted to test the correlation between GDP per capita and the pandemic. Same problem of data cleaning with the Anglo-Saxon representation. Once this was done I personally preferred to use nominal

Leave a Reply

Your email address will not be published.