Dimensional Data

Do you want your Delphi and C++ Builder Applications to connect with Access Database ? Do you need to manage some of the Access Database services such as creating, compacting database? How to start ? Don’t worry, FireDAC components offers robust components to connect with Access Database. FireDAC.Access Sample app demonstrates how to use FireDAC to work with access Database. You can find Delphi code samples in GitHub Repositories. Search by name into the samples repositories according to your RAD Studio version. Components used in the Sample App: TFDQuery : To execute SQL queries, browse the result sets, and edit the result set records. TFDPhysMSAccessDriverLink: To link the Microsoft Access driver to an application and set it up. In general, it is enough to only include the FireDAC.Phys.MSAcc unit into your application uses clause. It is used to specify access ODBC driver name and access the ODBC driver connection parameter common for all connections. TFDConnection : To establish a connection to a DBMS and to manage associated datasets. TFDAccessService: Class Implementing Microsoft Access database for creating, dropping, compacting, and repairing services. And some of the UI components, like TDBGrid,TDBComboBox, TFDGUIxWaitCursor1,TFDGUIxLoginDialog1,TFDGUIxErrorDialog1 Implementation Details: The simplest way to configure connection to MS Access database at run time is to build a temporary connection definition: In the sample, the temporary definition is created when the  item is selected in the Connection combo box.  Open the following database: C:UsersPublicDocumentsEmbarcaderoStudio20.0SamplesdataFDDemo.mdb. In the demo database, the Categories and Products tables have one-to-many relation by CategoryID field. Mention the query to the qryCategories.SQL property and qryProducts.SQL property. Finally, qryProducts.MasterSource property is set to dsCategories, while the MasterFields property is set to CategoryID. This creates a master-details relationship between the datasets. Simple queries execution is demonstrated via the ExecSQL method of TFDConnection. The management of databases, such as: creating, dropping, compacting/repairing, and setting a password is done using TFDMSAccessService component. This demo demonstrates how to create and compact the user database.  Check out the full article in the DocWiki about the FireDAC.Access Sample. FireDAC.Access Sample App Check out the full source code for the FireDAC.Access projects for Delphi and C++Builder over on GitHub.

Most of Delphi and C++ Builder developers utilize preinstalled components and libraries to compress and decompress files. For instance, the System.Zlib which supports gzip and deflate, the System.Zip is also helpful to handle .zip archive files. Moreover, Indy’s TIdCompressorZLib which is based on Zlib.  But that is not it. There are more different libraries based on different compression algorithms and more modern techniques, for instance, the Brotli – Brotli is similar in speed with deflate but offers more impenetrable compression. Brotli is open-sourced under the MIT License by Google. Brotli compressed files have .br extension. To connect your Delphi or C++ Builder VCL and FMX application with the Brotli library we can rely on Brotli Compress library from WINSOFT which offers to use the Brotli library easily. Brotli itself is free to use and distribute, but the Brotli Compress by WINSOFT is a commercial library and if you would like to use that library you should get a license! Uses Brotli library Supports Windows 32 and Windows 64 Available for Delphi/C++ Builder 6 – 10.4 After downloading the Brotli from WINSOFT you should configure the library into your RAD Studio. You can follow the tutorial here that shows the steps to configure without errors. Since the Brotli itself is a whole compression library your application should have brotlilib.dll – Dynamic-link library. You will get those files within the Brotli Compressor by WINSOFT in a Library Folder. Brotli Compressor Library has two main classes: TBrotliEncoder  TBrotliDecoder As you can observe the TBrotliEncoder encodes and TBrotliDecoder decodes the files with the given parameters. Furthermore, the OnProgress event provides the decompression and compression progress info. Additionally, you can set encoding quality with the Quality property. Here is the Brotli library demonstration video that shows the demo application in action.  These are demo projects’ UI: DEMO UI This is how you can encode with Brotli: InputStream := TFileStream.Create(EditFileName.Text, 0); try OutputStream := TFileStream.Create(ChangeFileExt(EditFileName.Text, ‘.br’), fmCreate); try with TBrotliEncoder.Create do try if RadioButtonGeneric.IsChecked then Mode := emGeneric else if RadioButtonText.IsChecked then Mode := emText else Mode := emFont; Quality := Round(TrackBarQuality.Value); OnProgress := Self.OnProgress; Compress(InputStream, OutputStream); finally Free; end; finally OutputStream.Free; end; finally InputStream.Free; end; 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 InputStream := TFileStream.Create(EditFileName.Text, 0);     try       OutputStream := TFileStream.Create(ChangeFileExt(EditFileName.Text, ‘.br’), fmCreate);       try         with TBrotliEncoder.Create do         try           if RadioButtonGeneric.IsChecked then             Mode := emGeneric           else if RadioButtonText.IsChecked then             Mode := emText           else             Mode := emFont;           Quality := Round(TrackBarQuality.Value);           OnProgress := Self.OnProgress;           Compress(InputStream, OutputStream);         finally           Free;         end;       finally         OutputStream.Free;       end;     finally       InputStream.Free;     end; Here is how you can decode with the Brotli: InputStream := TFileStream.Create(EditFileName.Text, 0); try if CheckBoxCheckIntegrity.IsChecked then OutputStream := nil else OutputStream := TFileStream.Create(ChangeFileExt(EditFileName.Text, ‘.uncompressed’), fmCreate); try with TBrotliDecoder.Create do try OnProgress := Self.OnProgress; Decompress(InputStream, OutputStream); finally Free; end; finally OutputStream.Free; end; finally InputStream.Free; end; 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 InputStream := TFileStream.Create(EditFileName.Text, 0);     try       if CheckBoxCheckIntegrity.IsChecked then         OutputStream := nil       else         OutputStream := TFileStream.Create(ChangeFileExt(EditFileName.Text, ‘.uncompressed’), fmCreate);       try         with TBrotliDecoder.Create do         try           OnProgress := Self.OnProgress;           Decompress(InputStream, OutputStream);         finally           Free;         end;       finally         OutputStream.Free;       end;     finally       InputStream.Free;     end; As you can see this is uncomplicated and you just need to implement the file selection […]

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.