Data analytic consistency of visual discoveries in sample datasets

1. A non-transitory computer-readable medium to store instructions, which when executed by a computer, cause the computer to perform operations comprising:

establish a connection from a data analytics application to an in-memory database via an in-memory database service;

initiate a query from a user interface of the data analytics application for automatic visual discovery in the in-memory database;

in response to the query, compute actual values of statistical signatures corresponding to analytical elements of a first sample dataset wherein the statistical signatures used consist of one or more of kurtosis, skewness or distinct count;

discretize the computed actual values by assigning bucket values to the computed actual values;

compute range values for the statistical signatures based on a step size, wherein the step size is automatically adjusted to determine maximum consistency across the first sample dataset and a second sample dataset;

compute an aggregate score based on a sum of the assigned bucket values corresponding to the analytical elements;

based on the computed aggregate score, rank the analytical elements of the first sample dataset; and

automatically display a menu of a consistent set of visual discoveries comprising a combination of analytical elements in the user interface of the data analytics application based on the ranking, wherein analytical elements from the consistent set are selectable to display corresponding graphical representations in an adjacent portion of a graph window.

2. The computer-readable medium of claim 1 , wherein discretizing the computed actual values further causes the computer to:

compute standard deviation and mean corresponding to the computed actual values of statistical signatures corresponding to the analytical elements;

compute the range values for the statistical signatures based on the computed standard deviation and the mean; and

assign bucket values to the computed range values for the statistical signatures.

3. The computer-readable medium of claim 2 , wherein discretizing the computed actual values further causes the computer to:

compare the computed actual values with the computed range values; and

based on comparison, determine bucket values for the computed actual values of statistical signature corresponding to the analytical elements.

4. The computer-readable medium of claim 3 , wherein computing aggregate score further causes the computer to:

rank the analytical elements of the sample dataset based on the computed sum of the bucket values, wherein the analytical elements with higher rank are identified for a combination of analytical elements.

5. The computer-readable medium of claim 4 , wherein ranking the analytical elements further causes the computer to:

identify the combination of analytical elements;

compute cumulative rank based on the individual rank of the analytical elements in the combination; and

display the combination of analytical elements with higher rank automatically in the user interface associated with automatic visual discoveries.

6. The computer-readable medium of claim 1 , wherein the analytical elements comprise dimensions and measures.

7. A computer-implemented method of improved consistency of visual discoveries, the method comprising:

establish a connection from a data analytics application to an in-memory database via an in-memory database service;

initiate a query from a user interface of the data analytics application for automatic visual discovery in the in-memory database;

in response to the query, computing actual values of statistical signatures corresponding to analytical elements of a first sample dataset wherein the statistical signatures used consist of one or more of kurtosis, skewness or distinct count;

discretizing the computed actual values by assigning bucket values to the computed actual values;

computing range values for the statistical signatures based on a step size, wherein the step size is automatically adjusted to determine maximum consistency across the first sample dataset and a second sample dataset;

computing an aggregate score based on a sum of the assigned bucket values corresponding to the analytical elements;

based on the computed aggregate score, ranking the analytical elements of the first sample dataset; and

automatically displaying a menu of a consistent set of visual discoveries comprising a combination of analytical elements in the user interface of the data analytics application based on the ranking, wherein analytical elements from the consistent set are selectable to display corresponding graphical representations in an adjacent portion of a graph window.

8. The method of claim 7 , wherein discretizing the computed actual values comprises:

computing standard deviation and mean corresponding to the computed actual values of statistical signatures corresponding to the analytical elements;

computing the range values for the statistical signatures based on the computed standard deviation and the mean; and

assigning bucket values to the computed range values for the statistical signatures.

9. The method of claim 8 , wherein discretizing the computed actual values further comprises:

comparing the computed actual values with the computed range values; and

based on comparison, determining bucket values for the computed actual values of statistical signature corresponding to the analytical elements.

10. The method of claim 8 , wherein computing aggregate score further comprises:

ranking the analytical elements of the sample dataset based on the computed sum of the bucket values, wherein the analytical elements with higher rank are identified for a combination of analytical elements.

11. The method of claim 10 , further comprises:

identifying the combination of analytical elements;

computing cumulative rank based on the individual rank of the analytical elements in the combination; and

displaying the combination of analytical elements with higher rank automatically in the user interface associated with automatic visual discoveries.

12. The method of claim 7 , wherein the analytical elements comprise dimensions and measures.

13. A computer system for improved consistency of visual discoveries, comprising:

a computer memory to store program code; and

a processor to execute the program code to:

establish a connection from a data analytics application to an in-memory database via an in-memory database service;

initiate a query from a user interface of the data analytics application for automatic visual discovery in the in-memory database;

in response to the query, compute actual values of statistical signatures corresponding to analytical elements of a first sample dataset wherein the statistical signatures used consist of one or more of kurtosis, skewness or distinct count;

discretize the computed actual values by assigning bucket values to the computed actual values;

compute range values for the statistical signatures based on a step

size, wherein the step size is automatically adjusted to determine maximum consistency across the first sample dataset and a second sample dataset;

compute an aggregate score based on a sum of the assigned bucket values corresponding to the analytical elements;

based on the computed aggregate score, rank the analytical elements of the first sample dataset; and

automatically display a menu of a consistent set of visual discoveries comprising a combination of analytical elements in the user interface of the data analytics application based on the ranking, wherein analytical elements from the consistent set are selectable to display corresponding graphical representations in an adjacent portion of a graph window.

14. The system of claim 13 , wherein discretizing the computed actual values comprises:

compute standard deviation and mean corresponding to the computed actual values of statistical signatures corresponding to the analytical elements;

compute the range values for the statistical signatures based on the computed standard deviation and the mean; and

assign bucket values to the computed range values for the statistical signatures.

15. The system of claim 14 , wherein discretizing the computed actual values further comprises:

compare the computed actual values with the computed range values; and

based on comparison, determine bucket values for the computed actual values of statistical signature corresponding to the analytical elements.

16. The system of claim 15 , wherein computing aggregate score further comprises:

rank the analytical elements of the sample dataset based on the computed sum of the bucket values, wherein the analytical elements with higher rank are identified for a combination of analytical elements.

17. The system of claim 16 , wherein ranking the analytical elements further causes the computer to:

identify the combination of analytical elements;

compute cumulative rank based on the individual rank of the analytical elements in the combination; and

display the combination of analytical elements with higher rank automatically in the user interface associated with automatic visual discoveries.