WEBVTT 00:01.850 --> 00:07.110 Hello everybody and welcome back to our first lesson about the none pay packets. 00:07.170 --> 00:12.980 Soon the last lessons we have seen that Python has limitations when it comes to large scale numerical 00:12.980 --> 00:19.710 operations python is simply too slow for this and therefore name the that package was created. 00:19.730 --> 00:28.130 And the roots go back to the year 1995 where it was developed under the name The MARICH And later in 00:28.130 --> 00:34.480 the year 2006 it was finally released as an unpaid package. 00:34.520 --> 00:35.930 Since then umpires. 00:35.960 --> 00:44.130 The first choice in Python when it comes to a numerical computations and fast vector US operations overlaps 00:44.240 --> 00:44.700 arrays. 00:44.720 --> 00:48.260 And we will see in the next lessons what exactly mean with this. 00:48.560 --> 00:56.150 However to give you an intuition typically so in many tasks that we use the NUM pay packets there is 00:56.180 --> 01:01.280 no further need for for loops or if and Elif and Ed statements. 01:01.280 --> 01:07.250 So with NUM PI we can reduce many lines of code and all of that can also increase the performance of 01:07.250 --> 01:13.760 our code and also umpires typically the number one choice for mathematical functions in general and 01:13.760 --> 01:18.930 in particular for linear algebra and also for a random number generation. 01:19.430 --> 01:19.790 All right. 01:19.790 --> 01:26.270 So let's stop coding and before we can use non PI we have to impart an umpire with the keyboard import 01:26.300 --> 01:29.480 num pi and then we say please use num pi. 01:29.510 --> 01:33.750 As with the previous Asian MP so that's by convention. 01:33.800 --> 01:37.590 So most of the users use the convention or the appreciation. 01:37.590 --> 01:42.180 MP fine umpire and then we create a list. 01:42.180 --> 01:51.750 So we already know lists a list from one to 10 including here at us and then we can transform the list 01:51.780 --> 01:59.670 into an so-called umpire Ray umpire race also India race fast and flexible container of large numerical 01:59.670 --> 02:06.480 data sets and these arrays enable us to perform operations on whole blocks of data and in a victorious 02:06.480 --> 02:07.110 fashion. 02:07.200 --> 02:10.350 So we will see later what I exactly mean with this. 02:10.500 --> 02:16.210 But first of all let's transform the listen to an umpire ravers the function p dot array. 02:18.000 --> 02:22.010 So now you can see that look similar to our lists here actually. 02:22.050 --> 02:27.100 And here with the array red you can see that here it's an umpire Ray. 02:27.330 --> 02:32.850 Earlier we learned that typically and to put a notebook and there's no need to use and the print function. 02:33.000 --> 02:40.530 So we can print the array by just executing the cell and it gives the same information then with the 02:40.530 --> 02:41.380 print function. 02:41.580 --> 02:46.530 However the way how the output is represented are formatted sometimes different and here. 02:46.530 --> 02:47.880 This is the case. 02:47.880 --> 02:51.030 So if we explicitly use the print function 02:55.760 --> 03:01.520 then we can see that the output is formatted in a different way and here the array looks like a vector. 03:01.530 --> 03:08.150 And yeah this is exactly how an num high array behaves later with operations. 03:08.190 --> 03:15.510 So like a mathematical back to also if you multiply a vector with a scalar let's say to then element 03:15.510 --> 03:18.740 twice each element will be multiplied by two. 03:18.810 --> 03:23.400 And that was exactly the behaviour that we didn't see the lists with lists. 03:23.460 --> 03:31.020 If we multiply a list by to be just get two times the copy of the list concatenated but we will go more 03:31.020 --> 03:33.240 into detail later in the next session. 03:33.630 --> 03:39.810 So now let's first check the data type of my array and it's num pie in the array. 03:40.200 --> 03:41.880 So umpire Ray or India right. 03:41.910 --> 03:48.330 I mean the same with it and same as lists the umpire Ray is a contain of elements so we can also iterate 03:48.390 --> 03:49.770 over an umpire Ray. 03:50.430 --> 03:57.980 So for element in my array I can print all elements 1 2 10 and 12. 03:57.990 --> 04:03.150 We have learned that lists can store many different elements so and for many different data types so 04:03.150 --> 04:13.510 our lists can star and enter into to float string and a boolean object so if we iterate over the elements 04:13.510 --> 04:20.590 of this list and print out the type of each element so we can see that erm began an integer and float 04:20.590 --> 04:22.250 and string and boolean object. 04:22.360 --> 04:29.470 So once an integer 2.5 is a float object dog is a string object and true is a boolean object. 04:29.470 --> 04:34.510 So now we are coming to the first and they are one of the most important differences between an umpire 04:34.510 --> 04:35.750 and the list. 04:35.770 --> 04:44.500 So let's try to transform the list here into an umpire Ray called a and then let's try to iterate over 04:44.500 --> 04:51.400 the elements and print out the type of each element and we can see each element as a string so we can 04:51.400 --> 04:58.630 see here also one is transformed into a string 2.5 is now a string and also true as the string and a 04:58.630 --> 05:01.060 string is typically the least common denominator. 05:01.060 --> 05:05.230 So yeah we can transform an into John a float into a string. 05:05.230 --> 05:11.440 You can also transform a boolean object into a string but you cannot transform a string into any of 05:11.600 --> 05:13.060 the other data types. 05:13.090 --> 05:16.500 So the thing is typically the lowest common denominator. 05:16.550 --> 05:17.720 Yeah and yeah. 05:17.760 --> 05:20.010 That's a general rule it out in an umpire Ray. 05:20.010 --> 05:26.560 You can only store one data type and typically and that's why if an umpire famous and successful listen 05:26.650 --> 05:31.840 that they are you typically use it with numerically the data sort with integers and floats. 05:32.240 --> 05:37.120 And that's one of the reasons why an umpire race are much faster when it comes to numerical operations 05:37.150 --> 05:41.840 because none by a dozen tests to check which data types it contains that knows. 05:41.840 --> 05:47.770 OK I only have integers off loads and I do not have to care about what data types I have. 05:47.770 --> 05:53.200 And yeah that makes it more powerful and more faster and then the numerical operations. 05:53.250 --> 05:53.700 All right. 05:53.710 --> 06:03.330 To give you a better intuition let's create an umpire Ray B with three integers so the type of still 06:03.330 --> 06:07.130 an umpire Ray and this the attribute detail. 06:07.140 --> 06:11.200 We can check the single data type that B contains. 06:11.220 --> 06:13.630 So here it's an integer. 06:13.720 --> 06:22.270 And now when we transform one element here into and float object by just adding a dot so he can see 06:22.270 --> 06:29.430 that num pi transforms all elements into a float object and let's check it here. 06:29.660 --> 06:32.990 The data type of the umpire AP is a float. 06:33.030 --> 06:36.440 Now if you transform two here into a string 06:39.630 --> 06:45.080 you can see that all elements are transformed into a string all right. 06:45.090 --> 06:50.880 So this is it for the time being and yeah we learned actually that an impact array contains only one 06:50.880 --> 06:55.050 data type and that's now one of the major differences to a list. 06:55.110 --> 06:58.250 And then the next lesson we learn about bacteria's operations. 06:58.260 --> 07:00.870 This is another main difference to a list. 07:00.870 --> 07:02.640 So we hope to see in the next lesson by.