>



+

-

Import data from outside sorce

After opening empty of data LeoDataAnalysis application the first step is add or insert data  series:

Step 1. Select source.

In the Control panel of LeoDataAnalysis do click on the tab: "Data" and then click on the button "Add" as it is shown on the picture:

Data tab

The dialog panel with available options for data import will appear:

Import data

Depending on source of your data check out the corresponding options from:

  • clipboard,
  • text file,
  • curve from image or pixels statistics,
  • MS Access database,
  • create empty series.
  • If you have in mind to type in data into created empty data series (table) you should type in all names of columns. An equivalent number of columns (fields) will be created. Later on you will be able to edit as number of columns as well their names if needed.

    If you already have at least one opened data series by this method you can add one more with exact the same number of parameters as in opened one.

    After selecting type of outside data source click at button "Next".

    Step 2. Import data series.

    From clipboard.

    If you use external application like a MS Excel for data preparing an import them into LeoDataAnalysis cannot be easier. Just select the region with data, and click "Copy". Then in LeoDataAnalysis choose "paste from clipboard" and click on "Next".

    If the first row on selected region contains non-numerical values LeoDataAnalysis will interpred them as names of parameters. Otherwise names of parameters will be assigned in style = a0, a1, a2 and so on.
    Note: The format of table in clipboard is based on "Tab" and "Space" separators between cells and "end of string" used as end of row. Standard sources of such data are spreadsheet like applications or Tables in MS Access or MS Word. Text file also can be copy-paste in LeoDataAnalysis.

    From text file.

    To import data from text file at first do chose the file by browsing your computer that contains data in form one record per line. Two example of  legitimate format of such file is presented below:

    Example 1 (first line contains names)
    x y
    1.05 3.4
    2.3 4.3
    5 1.2e1
    9 24.0
    Example 2: (only numeric fields)
    6, 4, 2.0
    3, 2.3, 1.0
    2e0, 1.2, 7.33
    43.1  ,1.e3, 33.6
    54, 222, 44.0
    77,555, 67.0
    99, 777.5, 88.0
    104, 1000.e-10,  88.0
    

    The first line can contains a names of columns; if it doesn´t the program will assign names in the style a0, a1, ... aN where N is number of the column.
    The actual dissection of the first line will be present in the appeared dialog panel for Example 2 like this:

    Read from text file

    This is actually a wrong situation as soon each field must contain only numerical letters divided by separator letters. By default space and "tab" sign are considered as separators. User can add any number of separators.  In presented case the commas are used as separators and as such must be typed in "Separators" edit control as it shown in following picture:

      Select fields in text file

    User also can by clicking on the corresponding lines do select fields only these of all existed to be read into LeoDataAnalysis.

    Click "OK" to read data.

    When a data source is a text file there is special option of automatically renewing data  that can be helpful in case of observation of gathering experimental results in real time. To do this:

    The example import data from text file spiral.txt will produce following picture when added as a first data series in LeoDataAnalysis:

    Spiral


    From image.

    One of the most  unique feature of LeoDataAnalysis is an ability extract and to incorporate data from an image. Scientists often have an impulse when reading article in professional journal to reanalyze data presented in form of chart. To get curves in numerical form is a challenging and let´s say the truth a very dull and time consuming job. LeoDataAnalysis can help to make in fast an accurate.

    Chose "Image" and click "Next" in "Import Data" dialog. The window that will give an ability to transfer visual data into numerical form will appear.

    Import your picture. There are two options to do this.

    1) From any image file in any of most popular formats. Click on the button  "From file". Browse and chose the file. All most common formats like bmp, gif, jpg are supported. As an example you can use the image of the picture:  temperature dependence of theoretical strength of ceramic materials .

    2) Paste image from clipboard. Most useful this option can be in case of internet published articles. Usually internet browsers permit to click on right button of the mouse when cursor is on the image evoking popup menu with element - "copy image". Choose it. Click on button "Paste from image from clipboard" in LeoDataAnalysis and get a picture. Of course any image from clipboard memory will do the job.

    Set physical coordinates on the chart: 1) Push down a button "Select chart area" and drag green rectangle expanding it in that way to cover a chart area with your data. Set minimum and maximum values of the rectangle area as it can be read from legend of the axes at the picture.

    By clicking by left button on the line or marks user is selecting records these later on will be imported in LeoDataAnalysis. "Alt +click" combination  will  remove the point pointed by cursor. After reaching satisfying result like shown on picture: Temperature dependence of theoretical strength
of ceramic materials

    click on "OK" button. The data series containing two parameters will be created in LeoDataAnalysis. For presented case the screen of LeoDataAnalysis will look like this:
    Temperature dependence of theoretical strength
of MgO

    In a case when one curve in the image are contrast enough and can be isolated, LeoDataAnalysis is able to set marks automatically. To do this at first press down button "Chose data area" and drag red rectangle to cover all this area. Press one of two buttons "Line" or "Marks".
    If a result of selection will be not satisfactory click on button "Remove points inside" and repeat the procedure with other setting of contrast sensibility.

    Create empty series.

    LeoDataAnalysis can at once create empty series with number and names of parameters reflected printed in the edit control.

    Later actual data have to be print out and all data edited as it is described in data manipulation page. General idea is to press "Show data" button at "Data" tab and manually type data one by one in the grid like control.

    Pixels color statistics.

    LeoDataAnalysis has the ability to read data directly from the image file. To do it go to Menu Data Read image file ⇒as RGB (or HSV)..

    The standard file search dialog will appear. Select any image file and open it. Depend on what color representation model you will chose histogram of colors distribution will be displayed. For example for the picture with basic red, green, blue color areas : distribution colors in RGB scheme will be like this: and for HSV scheme:

    HSV (Hue - direction from white color - the color itself, Saturation - intensity of color, Value - darkness) scheme of color representation has the advantage of separation colors in image permitting at least to count the number of brushes of palette in the image.

    The exact formula of calculating hue is quite complex. One can see that colors of spectra spread from red 0 to red on other side 255  (violet is 200, green 81 and blue 162). Saturation is calculated as an average deviation from value and normalized to interval 0-255. Value in its part is here an average of RGB values.

    Note and warning.. The image contains a lot of data. Due to representation of the data in LeoDataAnalysis a million of records that is usual for even average size image file will slow down a time of calculations especially for some of the statistical schemes. Just use common sense and be caution.

    From MS Access database file.

    LeoDataAnalysis can import data saved in database file. As soon there are numerous formats and variations of SQL syntactic of databases and connection protocols we can only claim compatibility with MS Access. In general we don´t recommend this type of data source if there is any alternative. Usually most database clients have option to save dataset in form of txt file.

    Nevertheless if you decide to use your database file as data source for LeoDataAnalysis the routine is following.
  • Browse your file system and open a database file. For example you can download and use as a test example the MS Access file couple.mdb
  • If the file will be recognized as a legitimate database file the "Database" dialog will be open:

      Select table in database file In the the dialog do select one table of the database that contains the data or use advantage of running your own sophisticated SQL statement to create dataset with special properties that could sometime perform a fancy job.


  • Press "OK".

    No matter if you select one of the tables or run SQL-statement next "Database" dialog will be displayed with available fields like this.

     Select fields in database table

    Important to know that only fields in numerical or text format containing only numerical letters will be read in LeoDataAnalysis. In case of breaking this rule for any cells all corrupted records will be skipped during reading. The same is valid for situation if you run SQL-statement. Click "OK".

    As a demonstration of ability of SQL statements to produce meaningful calculation lets solve small statistic-combinatorial problem:

    Let presume that we have a data of the age of bachelors males and females in table "couple" in our example couple.mdb database. The task is to find a distribution of all possible age distinctions for different gender couples.

    Solution:

    1) With the help of master of queries in original database application such as MS Access we will create two query: age_male and age_female  by running SQL-statements:

    SELECT age INTO age_female FROM couple WHERE gender=”f”

    SELECT age INTO age_male FROM couple WHERE gender=”m”

    2) Then run LeoDataAnalysis. Go to tab "Data"; click "Add"button; select "couple.mdb" file; type SQL-statement:

    SELECT age_male.age-age_female.age as d_age FROM age_male, age_female;

    click "OK"; select "d_age" field, click "OK"; go to "View" tab and click distribution button histogram button.
    The result image will be like this:

    Distribution of ages gap for couples

    It is worth to note that in the spite that the initial couple table contains only 49 records a result of running shown above SQL statements we will have all possible combinations of age variance in different gender couples - 600 cases at all.

    Comments:

    About Products Data analysis Crystal growth E-Vault Downloads Donate Contact Site Map © LeoKrut