WEBVTT 00:00.780 --> 00:07.230 In this video we will create an interactive histogram with bluntly and Instagram's are useful in cases 00:07.230 --> 00:13.170 where we want to examine the distribution of values and the one major application and finance is to 00:13.170 --> 00:16.230 analyze the distribution of returns. 00:16.230 --> 00:24.240 So we are still imparted the stocks data frame and with the percentage change method we can create the 00:24.270 --> 00:26.580 returns data frame. 00:26.670 --> 00:35.190 So here we have that daily returns for our six stocks and then we can create a histogram and we can 00:35.190 --> 00:41.700 simply do this with the eye plot method and we pass histogram to the kind parameter. 00:41.760 --> 00:42.530 So let's have a look. 00:48.140 --> 00:53.390 So here we have the interactive histogram with Plumlee and. 00:53.420 --> 00:59.650 For example we can select one single stock for example the Apple stock and TV effort. 00:59.660 --> 01:02.390 That distribution of daily returns. 01:02.390 --> 01:08.250 So for example he and the interval from 0 to 0 point two percent. 01:08.450 --> 01:11.980 We have one hundred fifty six occurrences. 01:12.050 --> 01:15.200 So here we have uh the absolute frequency. 01:15.800 --> 01:23.360 And it's actually no surprise uh that the majority of daily returns is around uh 0 percent here. 01:23.360 --> 01:27.380 But we have also some uh value you see on the tails. 01:27.530 --> 01:35.640 So we have one day where we have minus 10 percent roundabout and even uh one day with a minus twelve 01:35.650 --> 01:36.300 percent. 01:36.320 --> 01:39.710 And we have also quite positive days. 01:39.710 --> 01:43.570 So here returns with 8 percent 9 percent. 01:43.910 --> 01:51.650 And we can also compare this this abuse with another stock for example the Coca-Cola stock so we have 01:51.680 --> 01:55.220 both stocks and that seems like that. 01:55.250 --> 01:57.640 The Apple stock is more volatile. 01:58.010 --> 02:06.070 So there's more risk and the more frequency here in the tails and of course we could simply calculate 02:06.090 --> 02:12.850 the variance or the standard deviation to figure this out but uh we can also see this year graphically. 02:12.870 --> 02:17.870 So this is in many cases more intuitive. 02:17.890 --> 02:23.350 However there is one pitfall when comparing many histogram side distributions and politely and I think 02:23.350 --> 02:30.430 you can hardly see it that politely tries to scale and customize the bits of each distribution separately. 02:30.430 --> 02:36.260 So you might see here that in one apple been there actually two Coca-Cola bins. 02:36.520 --> 02:41.700 So the width of an apple in this actually twice the width of a Coca-Cola bin. 02:41.860 --> 02:43.500 And that makes it hard to compare. 02:43.510 --> 02:50.380 But we can set the bin dimensions for all distributions and we can do this by passing a topper to the 02:50.380 --> 02:52.220 bins parameter. 02:52.360 --> 02:59.470 So we have a bins perimeter and we can pass a topper with three values. 03:00.010 --> 03:03.740 So the first value marks the start. 03:03.730 --> 03:11.200 And let's have a look here on the left hand side so the start might be a maybe a minus 20 percent to 03:11.200 --> 03:18.950 get minus 15 percent to get you all values so we have a minus 15 percent. 03:19.490 --> 03:26.700 Then the second value is for the end and maybe we should to see our 10 percent 03:29.800 --> 03:33.830 and then the third value is actually the width of all bins. 03:34.000 --> 03:39.820 And uh here we could choose maybe uh 0.01 percent per bin 03:43.230 --> 03:48.530 and actually this scale and this dimension applies uh for all of this abuse. 03:48.540 --> 03:50.400 So let's run the site here 03:55.500 --> 04:06.990 and let's again have a look here only at uh the Apple stock and let's add also the Coca-Cola stock and 04:06.990 --> 04:12.180 all with the same bin size that is easier to compare both distributions. 04:12.330 --> 04:20.130 So we can see that the Coca-Cola stock has more frequency here in the middle and on the other hand side 04:20.130 --> 04:24.160 of the Apple stock has more frequency here in the States. 04:24.240 --> 04:30.020 So the Apple stock is obviously more risky than the Coca-Cola stock. 04:30.300 --> 04:34.340 And we can also see on the left hand side our y axis. 04:34.530 --> 04:39.050 And here we have absolute uh quantities of frequencies. 04:39.210 --> 04:47.520 And if we have distributions within a different amount of observations then it might be better to actually 04:47.530 --> 04:50.610 normalize um the y axis. 04:50.610 --> 04:54.750 And uh we can do this with repair meter his norm 05:02.290 --> 05:06.350 and we can either pass probability or percentage. 05:06.370 --> 05:08.800 So first of all let's try a probability 05:16.230 --> 05:21.290 and instead of getting the absolute quantities we get here the relative quantities. 05:21.330 --> 05:23.640 Um in terms of probabilities. 05:23.640 --> 05:25.990 So yes 6 percent 5 percent. 05:26.310 --> 05:34.230 And instead of getting your 0 point 0 6 and 0 point 0 5 you can also get uh the percentage terms low 05:34.230 --> 05:40.160 6 and 5 and we get this by passing a percent instead of probability 05:45.880 --> 05:47.200 and that's roughly the So 05:50.480 --> 05:54.680 and here we have now 6 percent 5 percent and so on. 05:54.680 --> 06:01.280 So this was a short introduction to the plotline histogram and I hope to see also in the next video 06:01.310 --> 06:01.610 by.