Analyse your data for randomness

Step 1: To create the “Randomness test report”, click “Dashboards”.

Step 2: Select the “Analysis” option to view all the reports.

Step 3:  Select the “Randomness test report” from the analysis option.

Note:  You can see the default report in your work area. Drag it to wherever you want it to be placed. Also, you can resize it by dragging the bottom right corner.

Step 4: Click “Edit” menu on the report header to edit your report’s data, appearance and social settings.

Step 5: To link data to the report, select the desired file from the left file tree and select columns from data.

Step 6:  In “Settings” click the reporting properties you want to analyse in your report. The properties include:

  • Median crossing: It is a non-parametric (i.e. does not assume an underlying distribution in the data) test for randomness. The data values are replaced by 0 if they are less than the median and 1 if they are greater than the median. If the data is random, then the number of times 0 is followed by 1 or 1 is followed by 0 and is approximately normally distributed. The test needs to apply the z-statistic from normal probability tables.
  • Turning point: It is a non-parametric (i.e. does not assume an underlying distribution in the data) test for randomness. The data values are assigned 1 if the value is between preceding and subsequent values (xi-1 < xi > xi+1 or xi-1 > xi < xi+1), otherwise they are assigned as 0. The number of times 1 appears (m*) is approximately normally distributed. The test needs to apply the z-statistic from normal probability tables.
  • Rank difference: It is a non-parametric (i.e. does not assume an underlying distribution in the data) test for randomness. The data values are replaced by their relative rank position starting at 1 for the lowest up to n for the largest. The sum of the absolute rank differences between successive ranks is normally distributed for large collections of random data. The test needs to apply the z-statistic from normal probability tables to the sum of absolute rank differences.
  • Auto correlation:: It is a parametric (i.e. assumes an underlying normal distribution in the data) test for randomness. The expected value and variance of the lag-one auto-correlation coefficient are normally distributed, if, the data is random. The test needs to apply the z-statistic from normal probability tables.

Step 7: After selecting the properties, format your report with “Formatting tab” which allows:

  • Title-font: the title font size, type and colour.
  • Title-Background: the title background colour.
  • Header font: the header font size, type and colour.
  • Body font: the body font size, type and colour.
  • Alignment: alignment of the text.

Edit the settings of your visualisation for sharing on social media

To edit the information that will accompany the report when sharing on social media (Facebook, Twitter, etc.), select the ‘Social’ tab.

Share your report

Share your dashboard with reports or a specific report using “Share” tab under “Dashboards”.