Data Analysis 2: Data Visualisation - Computerphile
Vložit
- čas přidán 24. 08. 2024
- Seeing is believing - Dr Mike Pound helps us understand how to turn our datapoints into Powerpoints. This is part 2 of the Data Analysis Learning Playlist: • Data Analysis with Dr ...
This Learning Playlist was designed by Dr Mercedes Torres-Torres & Dr Michael Pound of the University of Nottingham Computer Science Department. Find out more about Computer Science at Nottingham here: bit.ly/2IqwtNg
This series was made possible by sponsorship from by Google.
The ‘Chicken’ dataset was adapted from this dataset:
stat.ethz.ch/R...
/ computerphile
/ computer_phile
This video was filmed and edited by Sean Riley.
Computer Science at the University of Nottingham: bit.ly/nottsco...
Computerphile is a sister project to Brady Haran's Numberphile. More at www.bradyharan.com
Check out the full Data Analysis Learning Playlist: czcams.com/play/PLzH6n4zXuckpfMu_4Ff8E7Z1behQks5ba.html
Did you or can you make the Data sets and R code used in this series available some place ?
@@chadross3176 Yes, this would be really helpful!!
Please make the datasets used in this series available! It would really help us!
pro tip : watch series at Flixzone. Me and my gf have been using them for watching a lot of movies these days.
@Angelo Griffin yup, I have been using Flixzone for months myself =)
The chickens/eggs/diet example was wisely chosen, perfectly presented and definitely relevant. Well done, you certaily deserve your thumbs-up!
Is there any way I can download the dataset to practice?
ករថ
@@aaetyyjaaffgg188 ទូក
I came up with the optimal diet for chicken... unfortunately it only works for spherical chicken in a perfect vacuum.
Do you know the aerodynamics of the chicken too?
@@Ethelgiggle Of course not. That's why it only works in a vacuum.
Can you fix me the data? I'm researching for an utopian colony for harvest in orbit.
😂😂😂
Im really regretting not studying computer science in nottingham. All of those people are awesome especially Dr. Mike. To have a class with him would be nice for sure.
It's funny how you were talking about types of data, and then the commercial before your video said that the "tacos are 15% tastier"
15% more people said that they were tasty? or the average rating they gave rose by 15%??
or the cows for the beef got 15% fatter? idk
15:56 Your favorite sensationalist news outlet: *Diet A kills older chickens!*
Data Visualisation's impact in the conclusion is often overlooked or underated. I remember in experimental physics class how when we had to make the final visualisations of our measurements, specially in histograms, the different in width of the intervals made a huge difference in the overall look of the graph, without even touching any of the measurements. It's interesting how changing something completely "legal" and without obscuring any result you still could imply a different behavior
That's a bug not a feature. If the width or your histogram bars or the distance between them impacts your conclusion then it's a problem (or you're doing marketing, not science).
@@myothersoul1953 definitely. It was kinda marketing. Because to see that the experiment was done "properly", they expected a bell curve on our measurements' histogram. So when we did it, and it did not really look like a bell curve (even though in truth it was) a tweak on those things made a much more "pleasant" looking curve. It didn't change much of the conclusion, but it was just way more clear to the eye that it behaved as declared.
if you like data visualizations, I just started this channel where i'll visualize data on a variety of different topics like sports, gaming, politics, etc! Come check it out if you like this stuff...
The boxplot do not plot the maximum and minimum of the data. It actually plots the wiskers until the 3rd quartile (q3) + 1.95 * IQR or 1st quartile (q1) - 1.95 IQR. IQR means InterQuartile Range or the distance between q1 and q3.
I didn’t think I could like Dr. Mike any more than I already do, but then he said one of his favorite TV shows was Frasier and I was proven wrong.
Those down votes came from the pie chart enthusiasts.
I see Mike Pound, I click. What a good day
What makes a good graph? I can really recommend the book "The Visual Display of Quantitative Information" by Edward Tufte. In fact every scientist, politician and journalist should be mandated to read it in my opinion
i've started doing data visualization myself on my channel. I just started so the topics are a bit boring, but im about to move into sports, gaming, politics, pop culture, etc...come check it out if you want!
Thank you making the appropriate visualization clear. Excellent! People get duped easily by impressive visuals. Looks good, the scale becomes unimportant. Along comes a convincing presenter, and the truth gets lost. MS Excel's (and other spreadsheet applications) statistical functions are fantastic, but if you apply them inappropriately, your visualization may look and be colorful, but doesn't really have a meaning.
Came here for the data talk, got me some shared Frasier love! Great video
the hard part of this work is not the analysis, but the preparation. 9:1 of the work. the R/tidyverse is a good tool
In the last analysis, would it be appropriate to use the min/max of the data for B and C to limit the ages of the chickens for A? I.e., take only those chickens on diet A which were at least as old as the youngest chickens of B or C?
This might wind up reducing the data points available for A, but would remove the younger age bias.
This whole series is awesome. Well explained. I wish you did it before I wrote my thesis :D
This guy loves when things make sense
farmers hate him. see how he gets younger chickens with this amazing diet
Mike Pound is back!!
Currently writing out letters so people will hire me as a "data manager" listening to D A D D Y P O U N D gives me strength doing so. Ty ty
(this is a great resource genuinely ty)
Wait a second. That first graph, a time series, does make sense to geoscientists and climate change. I do agree there are too many curves but a legend should explain that.
Love it! Can you please link to your R-code?
A +1 as always, but especially for Frasier this time. :D
The chicken thing was extremely insightful and makes me question my own diet
Would that be too much if i ask for the upload of the sample data sheet?
What's with the dutch angle?
to distinguish theory from practice parts
For a novice to practice, would be great if you could make the Chicken.csv file available for download???
@15:00.. from 3 to 4 eggs are not a 20% increase but a 33% increase.. (3x1.20 = 3.6 and 3x1.33..= 4)
Could you explain why it makes sense to create plots about the age with the chicken dataset. It seems unintuitive to me when you measure the age of the same chicken over time and then create a histogram / boxplot.
@9:15 that’s not an assumption, that’s the intermediate value theorem.
Better than my lectures! And I am studying Computer science...
Might consider switching to Nottingham :D
If you're into data, i just started my channel where I'll do different data visualization mini-projects. The topics have started off basic, but ill get into gaming graphs and sports stats, politics, etc very soon. Come check it out!
good jab.
Excelent video!
Mike’s sweater is fresh. Where do I get one
Data analysts graduated in online courses: it is what it is.
What’s the price point between the diets and how much money per egg is being realised? Some more questions for charting
Pie charts are great if you want to see how much a data point takes up from the total
Love that little glimpse of R and ggplot. :)
Very nice analysis on the age, diet, eggs part. Thanks for the nice vid, though.
OK - I've finally subscribed
Is there a way to prove causation using stats? Or is only through lots of probes you can proof causation?
I like that you use ggplot2
Imho you do not necessarily need more chicken. you could look on the effect of age within each diet and later even check for interactions between observations e.g. age and diet. Although there you might miss some young chicken on B diet.
...feels odd to write things like that 😂
Hopefully I see that in later episodes. It has been a great series so far 🤗👍
Is there any way I can download the dataset to practice?
Amazing !!!!!
Awesome class! Thank u very much o /
To add a mathematicians view of the bar chart, we would say the labels == categorical data. In the previous video I think i would map the N of NOIR i.e. Nominal data to Categorical data. Would we say Nominal Data --> implies Categorical Data? Is the reverse true?
Sorry, more maths chatter on the histogram. The histogram 'labels' are usually called bins and will show the frequency of chickens on the y axis which fall into a range of continuous data. I guess in a computer science sence the range would probably (but not always) be a float rather than an integer. All this said, Dr. Mike Pound is wonderful - but just as he has bug bears with plotting subsets of data and wrong chart types, as a mathematician saying the frequency of chickens at 250 months is highest is not strictly correct because the bin it refers to is a range.
I had to look up the typical lifespan of a chicken, which is something I've never had pause to consider before.
Rich Fi so, how long it is?
Please have look again at the "bar chart for frequencies". you shouldn't use it to deduce the density or distribution. it's not a histogram unless you have a class width of one. it's the standardized frequency that you plot in a histogram. nice video though.
Dr. Mike given this data and assuming that perhaps this is the only data we had - if one were to make an operational change to their chicken business, would it not be logical to alter the diet of the older chickens in order to test the diet hypothesis? (because we can not make the older chickens any younger)
Are the R scripts and csv files that are used available for download anywhere?
4:01
The value of the USD plummeted, not of the JPY...
This is an amazing set of talks on data. But, perhaps the most amazing part is the old dot-matrix paper you found!!!! Do they still make that????
so, you're new to this channel, right?
Mike saying T H I C C in the first 10 seconds = insta like
I see Mike is an "=" man, the heathen. GREAT video, nonetheless!
Very interesting! Thanks :)
Bravo 🙌 🎉 👏 :)
This is a killer jumper, where can I buy it
On the last two graphs between diet B and C, the chickens on diet B appeared to be the youngest unlike the last box plot graph which shows that chickens on diet C are the youngest.
Someone noticed the same thing or it 's just me?
Here in Spain there are thousands of cases of 1:50 example :_(
Can you please share csv files?
6:55 remind me of japanese candlestick to read price.
That's why it's sometimes also called a "candlestick graph"
This video hit all the points ... R^2=1 🤣
omg where are the macs :(
Off topic - interesting chicken ages - the average modern chicken would be extremely lucky to get to 100 weeks old (my inner vegan is activated).
Frasier!!! 👍🏻👍🏻👍🏻
Why does the guy holding the camera have so much trouble holding it straight
9:35
Laughs in WSL.
Lindows?
Yves brought me here.
Can the chicken be named Bob loool
T A B L E A U
So Mike needed an excuse to learn R, ey? :D
I always though data analysis is something super high level stuff, but seems like it's just statistics?
Did you say OS „Ex“?? 😳😜
Great videos. Wish they were more concise. Hard to carve out time to watch them in series.
I feel like people pushing bitcoin do this all the tim...e XD
Says bar chart, draws column chart. Love the series though!
Column chart is just the Excel name for it, no? In most cases they'd be referred to as a bar charts or bar plots.
If you like data stuff like this, Ive started doing small data visualizations on my channel and i'll soon visualize stuff like sports, gaming, politics, etc! come check it out if you want to!
Use data.table, Mike :(
First
You have lots of acne on your face. That's data visualization.