WEBVTT 00:02.380 --> 00:08.730 Hello and welcome to the none pay assessment Crist's solution so same as with the pace and basic. 00:08.740 --> 00:13.510 We have 15 questions that cover all topics we need for our sticker. 00:13.600 --> 00:15.780 Case study and. 00:15.930 --> 00:21.070 Yeah if you have the impression that you need some more background on the basics feel free to watch 00:21.070 --> 00:27.400 the next few sessions about the basic none pay coding and follow up questions. 00:27.400 --> 00:31.570 We have the assumption that we already imparted an umpire. 00:32.500 --> 00:34.110 So let's start our quiz 00:37.810 --> 00:44.980 and the key message of the first question is that none pay array only contains one data type and by 00:44.980 --> 00:48.790 converting a list the different data types into a new pay array. 00:48.790 --> 00:52.060 All elements are converted to one data type. 00:52.240 --> 00:54.310 Actually to the lowest common denominator. 00:54.310 --> 00:57.760 And in this case that's the data type string. 00:57.760 --> 01:03.910 So with quotation marks we can transform an integer and will end and float to a string. 01:03.910 --> 01:06.580 So that correct answers string only 01:12.090 --> 01:19.050 here on this question we have an umpire ravers five elements and we have a so-called vector rise or 01:19.110 --> 01:23.670 element y's operation and where we add tend to each element. 01:24.210 --> 01:28.420 So actually we should get the eleven twelve thirteen fourteen fifteen. 01:28.420 --> 01:30.130 So this is correct yeah. 01:30.420 --> 01:37.080 And the key message is that in contrast to lists element y's operations are pretty easy here with an 01:37.080 --> 01:37.470 umpire 01:43.390 --> 01:48.790 so here we have two codes and they are we should determine how many of these codes the return uh the 01:48.790 --> 01:49.660 output. 01:49.690 --> 01:51.960 Six nine twelve 15 18. 01:52.210 --> 01:54.540 And let's have a look at the code one. 01:55.150 --> 02:02.140 So also here we have an element of ICE operation where we first add to and then multiply with three. 02:02.170 --> 02:12.250 So let's start with zero plus two is two times three is six then one plus two is three times three is 02:12.250 --> 02:24.440 nine two plus two is four times three is twelve three plus two is five times three is fifteen and four 02:24.440 --> 02:28.570 plus two is six times three is eighteen. 02:28.670 --> 02:36.580 So code 1 works and let's have a look at CO 2 Here we have the function P arrange and NPR range arranges 02:36.650 --> 02:44.600 integers and the range from six including to 19 excluding and only every third element. 02:44.630 --> 02:54.250 So we are starting with six then we have nine 12 15 and 18 so also Code 2 produces this output here. 02:54.320 --> 02:57.470 So the correct answer is both 03:03.420 --> 03:09.510 so here in this question we have a higher rate with 10 elements and we slicing from the next position 03:09.510 --> 03:17.410 for including to the next position eight excluding and also here the zero based indexing applies. 03:17.400 --> 03:25.440 So we start here with the 0 1 2 3 4. 03:25.620 --> 03:36.340 So our first element of our slices the 5 5 6 7 8 excluding. 03:36.360 --> 03:38.970 So our last element and our slices eight. 03:39.690 --> 03:47.290 So this should be the correct answer. 03:47.390 --> 03:55.640 Here we have some num pi array of us five elements 10 20 30 40 50 and we assign the arrival right a 03:56.960 --> 03:59.180 and then we make a slice of array. 03:59.180 --> 04:08.720 So we take them the first two elements and assign the variable array B so or a B contains of the elements 04:08.780 --> 04:18.740 10 and 20 and then we change the first element of array B and set it to 1 then actually so this answer 04:18.800 --> 04:24.730 is correct the first element of array B is null 1 and actually give us num pay raise then we create 04:24.730 --> 04:28.300 here a slice of array we are not creating a copy. 04:28.430 --> 04:36.410 So Arabia is referencing to the same original object but only to the first two elements and if you change 04:36.710 --> 04:43.580 one of the first two elements of array B we also change the elements of array A's so array a the first 04:43.610 --> 04:45.640 element of array a small one. 04:45.980 --> 04:51.690 And if you want to prevent this behavior of your we have to make a copy of the array a. 04:51.770 --> 04:58.340 So this is a common source for mistakes actually you make a slice and change some element and a slice 04:58.450 --> 05:05.450 of an none pay array and then also the original A changes and we might not even recognize that. 05:05.480 --> 05:06.920 So this is important here 05:09.780 --> 05:10.040 okay. 05:10.050 --> 05:10.560 The question. 05:10.560 --> 05:10.870 Yes. 05:10.880 --> 05:11.190 Yes. 05:11.200 --> 05:16.460 Which statement is incorrect so incorrect as then that the first element of array is still tense. 05:16.490 --> 05:17.610 Oh no that's correct 05:21.590 --> 05:30.080 here we are creating a one dimensional array with elements from 0 to 7 and then we reshape the one dimensional 05:30.080 --> 05:36.170 array to a two dimensional matrix with four rows and two columns. 05:36.170 --> 05:39.350 So here we have four rows and two columns. 05:39.350 --> 05:41.560 And this is the correct answer. 05:47.110 --> 05:51.760 Here we have a matrix a two dimensional array with a shape 3 5. 05:51.760 --> 05:55.340 So we have three rows and five columns. 05:55.340 --> 05:59.370 And yeah we're we want to index seven. 05:59.380 --> 06:07.350 So here this row is that next position the row and the second row is that index position 1. 06:07.390 --> 06:14.710 So let's move to the column here is column at index position 0 column meeting index position 1 and column 06:14.710 --> 06:16.240 at index position 2. 06:16.690 --> 06:29.380 So row at index position 1 and column and index position 2 should last for 7. 06:29.450 --> 06:36.770 Again we have the same matrix and as this kind of a tricky question so we slice the matrix we take all 06:36.770 --> 06:40.840 rows and the column at an exposition too. 06:40.850 --> 06:46.700 So this this this column here and then we add 100 to each element. 06:46.860 --> 06:53.100 And if you would print this line of code here then you would end up with this answer here but actually 06:53.100 --> 06:57.160 here with this code we are not changing the Matrix itself. 06:57.240 --> 07:04.010 So we are not assigning new values to the column at an exposition tours or actually we haven't changed 07:04.140 --> 07:05.730 the original matrix at all. 07:05.730 --> 07:11.660 And the correct answer is yeah the first one so the matrix my matrix is unchanged. 07:18.590 --> 07:26.600 Here we want to create 20 random integers in the in terms of 1 to 10 both and including and we want 07:26.600 --> 07:29.770 to select all elements that are greater than 5. 07:30.050 --> 07:38.050 And we set up the random C to 1 2 3 to ensure our reproducibility and let's have a look at the answers. 07:38.120 --> 07:42.910 So we want to create 20 random integers in the interval 1 to 10. 07:43.040 --> 07:49.130 So the starting point is including and the ending point and excluding and we need to have 20. 07:49.150 --> 07:54.670 So here this is correct and also this is correct and the second answer is not correct here. 07:54.950 --> 07:59.990 And now the second key message of this question is boolean indexing. 07:59.990 --> 08:08.090 So if you look here at this answer then for each of the 20 random integers we check whether the integer 08:08.090 --> 08:09.880 is greater than 5. 08:09.950 --> 08:16.910 And then we create here an NDA array with twenty billions and print here the end array with Julian's 08:17.880 --> 08:25.380 but actually we do not get here and the India test that's greater than 5 so he and the first answer 08:25.770 --> 08:32.580 we have are we in indexing so we select all elements of our array with 20 integers which are greater 08:32.580 --> 08:33.180 than 5 08:39.690 --> 08:42.810 here we calculate integers to the power of three of us. 08:42.810 --> 08:49.650 List comprehension and num pi code and the key message is that for large numerical operations like in 08:49.650 --> 08:53.700 this case here num pay us fifty two hundred times faster. 08:54.450 --> 09:02.510 So the fastest code should be here in the name pi code with only 1000 calculations and here for code 09:02.510 --> 09:06.030 number 1 and 2 with the same amount of calculation. 09:06.030 --> 09:10.750 So we have 100000 here the unpaid code should be much faster. 09:11.090 --> 09:19.160 So answer 3 to 1 is correct. 09:19.360 --> 09:25.810 Here we have an umpire array from 1 to twenty and we want to create a new array that contains all elements 09:25.930 --> 09:28.120 that fulfils two conditions. 09:28.210 --> 09:33.180 So first of all the element must be divisible by three without remainder. 09:33.730 --> 09:38.200 And the second condition is that the element three and must not be in the array. 09:38.200 --> 09:40.960 So we have two conditions and both must be met. 09:40.960 --> 09:42.610 So we need here and sign. 09:42.610 --> 09:45.260 Here we have an sign your F and or sign in here. 09:45.280 --> 09:48.420 And so this is not the correct answer. 09:48.550 --> 09:55.380 And what we need is really in slicing so this answer is incorrect as it returns only the boolean array 09:55.390 --> 09:58.860 is showing for each element of the array here. 09:59.080 --> 10:06.520 If both conditions are met but the answer does not return the elements itself that fulfill both conditions. 10:06.520 --> 10:08.190 Here we have two conditions. 10:08.200 --> 10:13.900 They are linked with an end and then we are creating as here a bully an array with truth or false for 10:13.900 --> 10:16.640 each element checking of each element. 10:16.670 --> 10:22.780 Fulfills both conditions and then we slice the array only for the elements that fulfill both conditions 10:22.790 --> 10:24.280 so this answer yes correct 10:27.780 --> 10:30.950 here we create 10 random normal distributed numbers. 10:30.960 --> 10:36.330 This mean 50 and standard deviation 10 and we calculate the mean. 10:36.480 --> 10:37.880 Over these 10 numbers. 10:37.920 --> 10:43.510 And we repeat this process 1000 times so we have 1000 simulations. 10:43.800 --> 10:48.360 And in the end we want to return an array with 1000 mean values. 10:48.420 --> 10:53.160 So the question is how many of the following codes are correct. 10:53.160 --> 10:54.180 And until. 10:54.180 --> 11:02.280 Here are three codes are identical so we create 10 times 1000 normal distributed numbers with mean 50 11:02.280 --> 11:04.430 and standard deviation 10. 11:04.560 --> 11:09.610 So we have 10000 the number distributed numbers and code 1. 11:09.690 --> 11:19.420 We reshape these 10000 numbers to 1000 rows and 10 columns and then for each row we take the mean over 11:19.420 --> 11:21.670 the 10 column so x equals 1. 11:22.510 --> 11:28.160 So here we should get a 1000 times the mean of 10 normal distributed numbers. 11:28.970 --> 11:38.510 And here in code to we reshape our 10000 numbers to 10 rows and 1000 columns and then we take the mean 11:38.570 --> 11:43.880 over all 1000 columns so 1000 times over 10 rows. 11:43.910 --> 11:51.170 So also here we should get an array with a 1000 the mean values and code 3 is actually identical to 11:51.200 --> 11:52.400 Kota 2. 11:52.580 --> 11:59.540 First of all we reshape the 10000 numbers into 1000 draws and ten columns and then we transpose the 11:59.540 --> 12:08.090 matrix into 10 rows and 1000 columns and then say mass and co 2 we take the mean for each of the 1000 12:08.090 --> 12:08.920 columns. 12:08.990 --> 12:21.320 So the answer is then that all three codes are correct. 12:21.400 --> 12:24.720 Here we have an array with seven elements from one to seven. 12:24.740 --> 12:30.080 And yeah gets code returns some the output far so we have fear in the middle of the four. 12:30.370 --> 12:35.450 And let's check what answers we have here so we have a person tired of the 19th person. 12:36.130 --> 12:42.860 We have the median and we have the sum so obviously the sum of their array cannot be for the person 12:42.860 --> 12:43.500 time. 12:43.520 --> 12:45.370 There are actually also not really. 12:45.470 --> 12:52.540 And the median is actually here the value where 50 percent of the other elements of the array is um 12:52.790 --> 12:55.530 less or lower than 50 percent is higher. 12:55.550 --> 13:02.800 So the fourth then the medium of our end the array and we can calculate the median with them the number 13:02.800 --> 13:04.480 of function and picked up medium 13:10.980 --> 13:17.160 so here we have a function Y equals cosine x plus except are we want to plot the graph and then Turnbull 13:17.160 --> 13:18.860 minus 10 to 10. 13:19.020 --> 13:22.230 And the question is which code best produces this graph. 13:22.230 --> 13:24.870 So let's have a look at our answers here. 13:26.510 --> 13:29.660 And to plot a graph you need how many data points. 13:29.660 --> 13:38.730 So NPL in space creates us 1000 evenly distributed data points in the interval minus 10 to 10 and MP 13:38.730 --> 13:43.570 range actually only creates 20 data points from minus 10 to 10. 13:43.610 --> 13:47.390 So this is not the optimal answer here. 13:47.390 --> 13:49.970 And comparing answer 2 and 3. 13:50.420 --> 13:54.370 So this in Python standard library no function like cosine. 13:54.380 --> 13:59.690 So we have to calculate the cosine with the number of function and picked up the coast. 13:59.780 --> 14:02.210 And this should be the correct answer here 14:06.910 --> 14:07.990 now our last question. 14:07.990 --> 14:14.290 So we have two dimensional data points that are represented by the ending arrays X for the x coordinate 14:14.950 --> 14:22.270 and y for the y coordinate and we want to fit a linear regression line that best fits the data points. 14:22.280 --> 14:28.180 Yeah the right function for it is the umpire function and P poorly fit where we pass them the arrays 14:28.190 --> 14:35.560 x and y and as we want to have a linear regression line our polynomial degrees should be 1 and not 2. 14:35.590 --> 14:38.770 So here we get a credit rhetoric regression. 14:38.770 --> 14:41.140 So the best answer here is the answer 1.