WEBVTT 00:01.190 --> 00:06.980 In this video we are going to further inspect our temperature times serious and we still have important 00:06.980 --> 00:15.040 pandas and we still have our data frame temperature with the daytime index here on the left hand side. 00:15.050 --> 00:22.610 And the two columns L.A. and New York and we can also call here the info method and we can see that 00:22.610 --> 00:26.370 in the column L.A. We have obviously two missing values. 00:26.480 --> 00:30.040 So we have thirty five thousand sixty four timestamps. 00:30.260 --> 00:38.320 And for L.A. we have only information regarding the temperature for thirty five thousand sixty two timestamps. 00:38.330 --> 00:43.910 So there might be two missing values and we can also further analyze our columns. 00:43.910 --> 00:45.710 This are the described method 00:48.930 --> 00:52.460 so we can see the amount of non missing values. 00:52.650 --> 00:55.140 Then here we have the mean temperature. 00:55.380 --> 00:59.950 So it's no surprise that for L.A. it's it's higher than for New York. 01:00.000 --> 01:02.490 Then we have the minimum temperature. 01:02.760 --> 01:06.860 And for New York it's a minus 22 degrees Celsius. 01:06.900 --> 01:09.720 And also here we have the maximum temperatures. 01:09.780 --> 01:16.890 And so the maximum for L.A. was higher than for New York with forty two point three. 01:17.040 --> 01:20.100 And we can also further analyze a single column. 01:20.130 --> 01:29.080 So for example the L.A. column and we can do this for example with the value columns method and we can 01:29.080 --> 01:34.440 see the most frequent value are the most frequent temperature was the sixteen point two. 01:34.990 --> 01:41.320 And actually one of the most powerful tools to analyze time see areas as graphically with line charts 01:42.190 --> 01:50.990 and therefore we import met plot lap top type plot S P T and by visualizing time series this line charts 01:51.020 --> 01:56.840 we can see how features like temperatures in New York and L.A. it changed over time and we can also 01:56.840 --> 02:04.400 compared to our mania time serious graphically so and in pain as we can do this quite simply with the 02:04.430 --> 02:05.330 plot method. 02:05.840 --> 02:11.680 So if you have a U.S. city plot method on our data frame temperature and the use of the fixed size 15 02:11.690 --> 02:12.320 7 02:16.170 --> 02:22.680 and 2 we can actually see in blue the time series for L.A. and an orange so the time series for New 02:22.680 --> 02:30.330 York and by default when we are calling the plot method on a data frame then pan us creates here a line 02:30.330 --> 02:31.220 plot. 02:31.380 --> 02:36.390 So you have a line plot with on the x axis the index. 02:36.390 --> 02:38.160 And here is a daytime index. 02:38.160 --> 02:40.180 We are here on the x axis. 02:40.190 --> 02:45.540 Actually our date time and on the y axis we have the temperature. 02:45.540 --> 02:52.670 So now here we have said the graph for New York and L.A. in one graph but we can also create subplots 02:52.680 --> 02:57.000 and we can do this here by passing true to the subplots per meter 03:02.240 --> 03:07.820 so here we have now the two subplots organized vertically one above at the other but we can also change 03:07.910 --> 03:15.530 this here and we've actually had the perimeter laid out so let's try this out here and then let's have 03:15.530 --> 03:15.860 a look. 03:23.270 --> 03:24.830 And to the perimeter lay out. 03:24.830 --> 03:27.970 We can pass a top were there on the left hand side. 03:27.980 --> 03:34.330 We can define the amount of rows and on the right hand side the amount of columns slowing. 03:34.370 --> 03:37.930 The default setting is that we have two rows and one column. 03:37.970 --> 03:44.510 But if you want to organize your our two subplots horizontally side by side we can also change this 03:44.510 --> 03:44.960 here. 03:46.020 --> 03:50.760 So we can say that if you want to f one row and two columns 03:53.380 --> 03:54.400 and here we are. 03:54.580 --> 03:58.720 So here we have organized the two subplots the side by side. 03:58.720 --> 04:04.570 And still it's quite hard to compare the two time series because the uh the scale on the BI axis is 04:04.720 --> 04:05.440 different. 04:05.950 --> 04:12.790 So here in New York we're starting with the minus 20 degrees and here for a myth about zero. 04:13.240 --> 04:17.690 And we've already learned that this additional apparent meter called shall buy it. 04:18.430 --> 04:27.220 So let's do this here and here we can define whether the two subplots should share the same y axis. 04:27.220 --> 04:31.030 And in this case you want to compare both Time series. 04:31.030 --> 04:37.300 So actually they should have the same y axis and therefore we can set here the issue by a power meter 04:37.320 --> 04:38.020 to a true 04:41.690 --> 04:49.080 and actually B I feel the same y axis for both subplots and we can see that in summertime there's no 04:49.080 --> 04:51.770 a huge difference between New York and L.A.. 04:51.800 --> 04:54.430 So here we are about 30 degrees. 04:54.710 --> 04:57.880 However there's a huge difference here in wintertime. 04:57.890 --> 05:03.620 So in winter time in New York we have to minus 10 minus 20 degrees. 05:03.620 --> 05:11.110 And here now a we hardly get temperatures and zero degrees here and still we can see here that these 05:11.110 --> 05:17.710 are not optimal grass because here we have so many data points that because we have our early data and 05:17.710 --> 05:25.010 it might make more sense to ask your flight sample a weekly or monthly data so far visualizing our time 05:25.010 --> 05:28.270 series it might be better to transform to weekly or monthly. 05:28.690 --> 05:34.660 And in the next videos you will also learn how to aggregate time serious to let's say weekly or monthly 05:34.660 --> 05:35.370 data. 05:35.740 --> 05:41.110 But for the time being we are now finishing with this video and I hope to see you on the next one by.