Dimensional Data

Are you looking for a powerful machine learning library? Try OpenCV library for Python. You can run it with Python for Delphi (P4D). P4D is a free and simple with which you can run Python scripts as well as create new Python modules and types in Delphi. Use Delphi and C++Builder and Python4Delphi to run Python scripts in  Python GUI apps for Windows. First, run Demo1 project for executing Python script in Python for Delphi. Then load script in text field and press Execute button to see the result. Go to GitHub to download Demo1 source. procedure TForm1.Button1Click(Sender: TObject); begin PythonEngine1.ExecStrings( Memo1.Lines ); end; procedure TForm1.Button1Click(Sender: TObject); begin PythonEngine1.ExecStrings( Memo1.Lines ); end; OpenCV is an open-source library for computer vision and machine learning that supports various programming languages including Python. With this library, you can do a lot of difficult operations, such as image processing, video analysis, feature detection, machine learning, computational photography, object detection. K-Nearest Neighbour In this example, we will consider solving the problem of finding nearest neighbors using OpenCV library. First, let’s randomly create 20 red points (family 0) and 20 green points (family 1). Then add 5 blue points. Using function train(), we will train the neural network. Function findNearest() returns k nearest neighbours (in our example k=3) for each blue point. It also calculates the distance to each found neighbor and determines the family of points from which more neighbors are found. import cv2 import numpy as np import matplotlib.pyplot as plt # Feature set containing (x,y) values of 20 training data trainData = np.random.randint(0,100,(20,2)).astype(np.float32) # Labels each one either Red or Green with numbers 0 and 1 responses = np.random.randint(0,2,(20,1)).astype(np.float32) # Take Red points and plot them red = trainData[responses.ravel()==0] plt.scatter(red[:,0],red[:,1],50,’r’,’s’) # Take Green points and plot them green = trainData[responses.ravel()==1] plt.scatter(green[:,0],green[:,1],50,’g’,’^’) # 5 new points newpoints = np.random.randint(0,100,(5,2)).astype(np.float32) plt.scatter(newpoints[:,0],newpoints[:,1],50,’b’,’o’) knn = cv2.ml.KNearest_create() knn.train(trainData,cv2.ml.ROW_SAMPLE,responses) ret, results,neighbours,dist = knn.findNearest(newpoints, 3) print(“result: “, results,”n”) print(“neighbours: “, neighbours,”n”) plt.show() 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 import cv2 import numpy as np import matplotlib.pyplot as plt   # Feature set containing (x,y) values of 20 training data trainData = np.random.randint(0,100,(20,2)).astype(np.float32)   # Labels each one either Red or Green with numbers 0 and 1 responses = np.random.randint(0,2,(20,1)).astype(np.float32)   # Take Red points and plot them red = trainData[responses.ravel()==0] plt.scatter(red[:,0],red[:,1],50,‘r’,‘s’)   # Take Green points and plot them green = trainData[responses.ravel()==1] plt.scatter(green[:,0],green[:,1],50,‘g’,‘^’)   # 5 new points newpoints = np.random.randint(0,100,(5,2)).astype(np.float32) plt.scatter(newpoints[:,0],newpoints[:,1],50,‘b’,‘o’) knn = cv2.ml.KNearest_create() knn.train(trainData,cv2.ml.ROW_SAMPLE,responses) ret, results,neighbours,dist = knn.findNearest(newpoints, 3)   print(“result: “, results,“n”) print(“neighbours: “, neighbours,“n”)   plt.show() Perspective Transformation of an Image To perform perspective transformation with an image use warpPerspective() function. The parameters of this function are the original image, the transformation matrix, and the size of the output image. Use getPerspectiveTransform() function to get the transformation matrix. You need to pass four points of the input image and the corresponding four points of the output image to this function. It is important, that three of the four points should not be on the same straight line. import cv2 import numpy as np import matplotlib.pyplot as plt image_path = “E:faces.JPEG” img = cv2.imread(image_path) pts1 = np.float32([[900,100],[1200,100],[900,400],[1200,400]]) pts2 = np.float32([[0,0],[400,0],[0,400],[400,400]]) M = cv2.getPerspectiveTransform(pts1,pts2) dst = […]

If you need to perform complex transformations or mathematical calculations with matrices or arrays, then Python Numpy library is exactly what you need. You can easy run this library with Python4Delphi (P4D). Python4Delphi is a free tool with which you can work with Python scripts and objects in the Windows GUI. In this post, we will look at how to run Numpy library with P4D. Now you can build Python GUI apps for Windows using a lot of Python libraries with Delphi and C++Builder and Python4Delphi. Just open and run Demo1 project. Then paste the Python script into the text field, press Execute button and get the result. Go to GitHub to download Demo1. procedure TForm1.Button1Click(Sender: TObject); begin PythonEngine1.ExecStrings( Memo1.Lines ); end; procedure TForm1.Button1Click(Sender: TObject); begin PythonEngine1.ExecStrings( Memo1.Lines ); end; Numpy library allows you to create multidimensional arrays and matrices and work with their properties. It also contains various functions for processing arrays and matrices. Let’s look at some simple examples of working with Numpy. If you have trouble compiling the Python you may need to have Numpy 1.19.3 installed. Create a matrix and get some properties This example shows how to create a 3-dimensional array and fill it with numbers from 0 to 29. Then, using the properties of this array, we can find out its shape, dimension, data type, number of elements. import numpy as np a = np.arange(30).reshape(2, 3, 5) print(a) print(a.shape) print(a.ndim) print(a.dtype.name) print(a.itemsize) print(a.size) import numpy as np a = np.arange(30).reshape(2, 3, 5) print(a) print(a.shape) print(a.ndim) print(a.dtype.name) print(a.itemsize) print(a.size) Basic operations with arrays Let’s take a look at the simplest conversions you can perform on arrays. Using concatenate() function, you can combine the values of two arrays into one. With function sort() you can sort ascending the values in an array. Function reshape() allows you to change the dimension of the array. import numpy as np arr = np.array([7, 10, 3, 11, 29, 15, 18]) print(np.sort(arr)) a = np.array([1, 2, 3, 4, 5, 6]) b = np.array([7, 8, 9, 10, 11, 12]) print(np.concatenate((a, b))) c = a.reshape(3, 2) print(c) import numpy as np arr = np.array([7, 10, 3, 11, 29, 15, 18]) print(np.sort(arr)) a = np.array([1, 2, 3, 4, 5, 6]) b = np.array([7, 8, 9, 10, 11, 12]) print(np.concatenate((a, b))) c = a.reshape(3, 2) print(c) Mathematical operations with matrix Function default_rng()  allows you to fill a matrix with random values. You can use integer or real numbers. In this example, we fill the matrix with integer values. Then we find the maximum and minimum element, the sum of all the elements in the matrix. It is also shown how you can multiply a matrix by a number and sum two matrices with the same dimension. import numpy as np from numpy.random import default_rng rng = default_rng() arr=rng.integers(20, size=(2, 4)) print(arr) print(arr.max()) print(arr.min()) print(arr.sum()) print(arr*2) arr2=rng.integers(5, size=(2, 4)) print(arr2) print(arr+arr2) import numpy as np from numpy.random import default_rng rng = default_rng() arr=rng.integers(20, size=(2, 4)) print(arr) print(arr.max()) print(arr.min()) print(arr.sum()) print(arr*2) arr2=rng.integers(5, size=(2, 4)) print(arr2) print(arr+arr2) And this is only a small part of what Numpy library allows you to do. Download Numpy and check how many possibilities it opens for your applications. Install Python4Delphi for building Python GUIs for Windows using Delphi easily.

In the Clang-enhanced C++ compilers, two consecutive right angle brackets no longer generate an error, and these constructions are treated according to the C++11 standard. C++03’s parser defines “>>” as the right shift operator or stream extraction operator in all cases. However, with nested template declarations, there is a tendency for the programmer to neglect to place a space between the two right angle brackets, thus causing a compiler syntax error. C++11 improves the specification of the parser so that multiple right angle brackets will be interpreted as closing the template argument list where it is reasonable. Right angle brackets example #pragma hdrstop #pragma argsused #include #include #include typedef std::vector > Table; // OK typedef std::vector> Flags; // OK int _tmain(int argc, _TCHAR* argv[]) { return 0; } #pragma hdrstop #pragma argsused   #include #include   #include typedef std::vector<std::vector<int> > Table;  // OK typedef std::vector<std::vector<bool>> Flags;  // OK   int _tmain(int argc, _TCHAR* argv[]) {           return 0; } Head over and find all of the different C++ language features now available in C++Builder. 

In this CodeRage session, you can learn how to speed up FireMonkey layout construction and improve the painting performance of FireMonkey applications. With Delphi Visual Designer you can design any user interface by easily dragging and dropping the components. If you need some customizations you can alter the properties of the components or just change the style. But complex layouts tend to contain more controls than necessary. TFmxObject – Base Class for FireMonkey Components Heavyweight class Several lazily initialized lists on top of TComponent ones Complex layouts need a substantial amount of time to construct, align and paint Performance penalties can be significant on mobile platforms So, how we can optimize our layouts then? Here are the possible layout optimization options that you can implement: Simplify layout designs Optimize images for screen size and density Avoid shadows, effects, or any unnecessary effects Implement Lazy loading Optimize layout hierarchies without changing the visual appearance of the layout Be sure to watch the session to learn and grasp real-world experience in optimizing FireMonkey applications!

WinRT is the latest application architecture from Microsoft for Windows 10 and components development. This webinar will show you how to take advantage of the new WinRT APIs and components in your Delphi and C++Builder applications. WinRT is a modern C++17 language projection for WinRT and it’s implemented entirely in header files. This leads to great performance, both in time and space, typically performing better and producing smaller binaries than any other language option for the Windows Runtime. In one sentence, WinRT is for the new world. Agenda The Windows architecture strategy What is WinRT? Why should we care about it? WinRT/UWP vs Win32 vs COM vs .NET The Ever Moving Strategy with Windows WinRT Components in RAD Studio Q&A Windows Runtime  Combine what’s great in COM and .NET Binary interface for language interoperability Metadata management is similar to .NET New COM as a natural part of all languages/platforms .NET, C++, JavaScript, Delphi and C++ Builder Build a new class library on the new runtime Cleanup the library compared to .NET Do not enforce managed code