Mathematical Views

Have you ever wondered what you would see if you were sitting at the origin and looking down at some spheres distributed on a logarithmic curve?

Thought not.

But why not take a look anyway? I've made a new page on Cats in Drag which lets you view this and other functions. If you want to suggest any other functions, please comment here and I'll get right on it.

"Uncanny" Royal Coincidence

A news story on the BBC Website caught my eye this morning. It reported the following "uncanny coincidence" between a Northern Irish baby and a Royal baby:

But both new mothers share the name Catherine, the same birthday - 9 January - and now their sons also share the same birth date.

~~ BBC News - Another Kate welcomes her own 'royal' baby

I decided to work out just how uncanny this is.

The Office for National Statistics states that 729,674 babies are born every year in the UK. This works out at 1,999 babies born each day, assuming that births are uniformly distributed, so there will be approximately 1,998 babies who share Price Nameless's birthday.

So, what is the chance of the mother of one of these babies having the same birthday as Princess Kate? To work this out I used a method similar to that which is used in the birthday "paradox", which tells us that in a group of 23 people there is a more than 50% chance of two people sharing a birthday, but that's another story.

First, we look at one of our 1,998 mothers. The chance that she shares Princess Kate's birthday is 1/365 (ignoring leap days). The chance that she does not share Princess Kate's Birthday is 364/365.

Next we work out the probability that none of our 1,998 mothers shares Princess Kate's birthday. As our mothers' birthdays are independent we can multiply the probabilities together to do this (this is why we are looking at the probability of not sharing a birthday instead of sharing a birthday). Our probability therefore is (364/365)¹⁹⁹⁸ = 0.00416314317.

Back to the original question, we wanted to know the probability that one of our mothers shares Princess Kate's birthday. To calculate this we do take 0.00416314317 away from 1. This gives 0.99583685682 or 99.6%.

There is a 99.6% chance that there is a resident of the UK who shares the same birthday as Princess Kate and had a child on the same day.

Uncanny.

But let's be fair. The mother in our story is also called Kate. So what are the chances of that? In fact, the same method can be followed, working with the probability of having neither the same birthday or name as Princess Kate.

I think it is safe to assume that this would still be considered news-worthy if our non-princess was called Katie, Cate, Cathryn, Katie-Rose or any other name which is commonly shortened to Kate, so I included a number of variations and used this fantastic tool to find the probability of a mother being called Kate. The data only goes back to 1996, but as the name is dropping in popularity, we can assume that before 1996 at least 1.5% of babies were called Kate. Disregarding males, we can estimate that 3% of mothers are called Kate.

If anyone would like the details of the rest of the calculation, please comment on this post and I will include it here. For anyone who trusts me and isn't curious, I eventually found that the probability of none of our 1,998 mothers share the same name and birthday as Princess Kate is 0.84855028964. So the probability of another Kate having a child on the same day and sharing Princess Kate's birthday is 0.15144971035 or 15.1%. Just over one in seven.

So this is as uncanny as anything else which has a probability of one in seven, such as the Royal baby being born on a Monday (uncanny!).

Raspberry Pi Weather Station

On 19 June, the USB temperature sensor I ordered from Amazon arrived. This sensor is now hooked up to my Raspberry Pi, which is taking the temperature every 10 minutes, drawing graphs, then uploading them here. Here is a brief outline of how I set this up (I'm writing this nearly two weeks after I did it so if you can't follow any of the steps let me know and I'll try to work out what I've missed out):

Reading the Temperature

I found this code and adapted it to write the date, time and temperature to a text file. I then set cron to run this every 10 minutes. It writes the data to a text file (/var/www/temperature2) in this format:

2013 06 20 03 50,16.445019 2013 06 20 04 00,16.187843 2013 06 20 04 10,16.187843 2013 06 20 04 20,16.187843

Plotting the Graphs

I found a guide somewhere on the internet about how to draw graphs with Python using Pylab/Matplotlib. If you have any idea where this could be, comment below and I'll put a link here.

In the end my code looked like this:

import time import matplotlib as mpl mpl.use('Agg') import matplotlib.pylab as plt import matplotlib.dates as mdates ts = time.time() import datetime now=datetime.datetime.fromtimestamp(ts) st = mdates.date2num(datetime.datetime(int(float(now.strftime('%Y'))),int(float(now.strftime('%m'))),int(float(now.strftime('%d'))),0,0,0)) weekst = st-int(float(datetime.datetime.fromtimestamp(ts).strftime('%w'))) f=file('/var/www/temperature2','r') t=[] s=[] tt=[] ss=[] u=[] v=[] g=[] h=[] i=[] weekt=[] weektt=[] weeks=[] weekss=[] mini=1000 maxi=0 cur=-1 datC=0 for line in f: fL=line.split(",") fL[0]=fL[0].split(" ") if cur==-1: cur=mdates.date2num(datetime.datetime(int(float(fL[0][0])),int(float(fL[0][1])),int(float(fL[0][2])),0,0,0)) datC=mdates.date2num(datetime.datetime(int(float(fL[0][0])),int(float(fL[0][1])),int(float(fL[0][2])),int(float(fL[0][3])),int(float(fL[0][4])),0)) u.append(datC) v.append(fL[1]) if datC>=st and datC<=st+1: t.append(datC) s.append(fL[1]) if datC>=st-1 and datC<=st: tt.append(datC+1) ss.append(fL[1]) if datC>=weekst and datC<=weekst+7: weekt.append(datC) weeks.append(fL[1]) if datC>=weekst-7 and datC<=weekst: weektt.append(datC+7) weekss.append(fL[1]) if datC>cur+1: g.append(cur) h.append(mini) i.append(maxi) mini=1000 maxi=0 cur=mdates.date2num(datetime.datetime(int(float(fL[0][0])),int(float(fL[0][1])),int(float(fL[0][2])),0,0,0)) mini=min(float(fL[1]),mini) maxi=max(float(fL[1]),maxi) g.append(cur) h.append(mini) i.append(maxi) plt.plot_date(x=t,y=s,fmt="r-") plt.plot_date(x=tt,y=ss,fmt="g-") plt.xlabel("Time") plt.ylabel("Temperature ($^\circ$C)") plt.title("Daily") plt.legend(['today','yesterday'], loc='upper left',prop={'size':8}) plt.grid(True) plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%H:%M')) plt.gcf().subplots_adjust(bottom=0.15,right=0.99) labels = plt.gca().get_xticklabels() plt.setp(labels,rotation=90,fontsize=10) plt.xlim(st,st+1) plt.savefig('/var/www/tempr/tg1p.png') plt.clf() plt.plot_date(x=weekt,y=weeks,fmt="r-") plt.plot_date(x=weektt,y=weekss,fmt="g-") plt.xlabel("Day") plt.ylabel("Temperature ($^\circ$C)") plt.title("Weekly") plt.legend(['this week','last week'], loc='upper left',prop={'size':8}) plt.grid(True) plt.gca().xaxis.set_major_formatter(mdates.DateFormatter(' %A')) plt.gcf().subplots_adjust(bottom=0.15,right=0.99) labels = plt.gca().get_xticklabels() plt.setp(labels,rotation=0,fontsize=10) plt.xlim(weekst,weekst+7) plt.savefig('/var/www/tempr/tg2p.png') plt.clf() plt.plot_date(x=u,y=v,fmt="r-") plt.xlabel("Date & Time") plt.ylabel("Temperature ($^\circ$C)") plt.title("Forever") plt.grid(True) plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%d/%m/%y %H:%M')) plt.gcf().subplots_adjust(bottom=0.25,right=0.99) labels = plt.gca().get_xticklabels() plt.setp(labels,rotation=90,fontsize=8) plt.savefig('/var/www/tempr/tg4p.png') plt.clf() plt.plot_date(x=g,y=h,fmt="b-") plt.plot_date(x=g,y=i,fmt="r-") plt.xlabel("Date") plt.ylabel("Temperature ($^\circ$C)") plt.title("Forever") plt.legend(['minimum','maximum'], loc='upper left',prop={'size':8}) plt.grid(True) plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%d %b')) plt.gcf().subplots_adjust(bottom=0.15,right=0.99) labels = plt.gca().get_xticklabels() plt.setp(labels,rotation=90,fontsize=8) plt.savefig('/var/www/tempr/tg3p.png')

If there's anything in there you don't understand, comment below and I'll try to fill in the gaps.

Uploading the Graphs to catsindrag.co.uk

Finally, I wrote a php script to upload the graphs to catsindrag.co.uk/weather:

$filename=Array("tg1p.png","tg2p.png","tg3p.png","tg4p.png"); for($i=0;$i<count($filename);$i++){ $localfile="/var/www/tempr/".$filename[$i]; $fp = fopen($localfile, "r"); $ch = curl_init(); curl_setopt($ch, CURLOPT_URL, "ftp://USERNAME:PASSWORD@catsindrag.co.uk/sites/www.catsindrag.co.uk/public_html/weather/".$filename[$i]); curl_setopt($ch, CURLOPT_UPLOAD, 1); curl_setopt($ch, CURLOPT_INFILE, $fp); curl_setopt($ch, CURLOPT_INFILESIZE, filesize($localfile)); curl_exec($ch); curl_close($ch); }

I hope this was vaguely interesting/useful. I'll try to add more details and updates over time. If you are building something similar, please let me know in the comments; I'd love to see what everyone else is up to.