topic Re: Life events as a percentage of births in DAX Commands and Tips

Life events as a percentage of births

Anonymous — Tue, 17 Nov 2020 21:38:49 GMT

I am building a Power BI dashboard off an Excel document that mainly utilizes pivot tables. In a pivot table, I am able to create a calculation that identifies the proportion of all births that have a life event, i.e. what proportion of all the births for a specific time period (month of birth) had a pneumonia event?

The Pivot Table creates these values using the fields of EVENT and BirthDate (summarized as a birth month BMON). In the Pivot Table the column are assigned to EVENT, and the rows to BMON. The summation of values is the COUNT of ID (the unique individual reference number), with the Show Values As option set to "% OF", Base field = EVENT and Base item set as 1_BORN. The resulting values are then displayed as percentages. Of all the individuals born in the month of Feb 2019, 3.5% encountered a pneumonia event is an example result.

The data is structured with fields for ID, BDAT, BMON, EVENT, AGE, and EDAT (reference ID, birthdate, birth month, EVENT, age at event and event date). There may be multiple rows for each individual depending on the events they encountered.

In BI, I am unable to replicate this Excel calculation.

Thank you very much for any insights!

ID BDAT EVENT AGE EDAT BMON

72281	7/29/2019	1_BORN	0	7/29/2019	7/1/2019
72281	7/29/2019	WEANED	69	10/6/2019	7/1/2019
115188	11/16/2019	1_BORN	0	11/16/2019	11/1/2019
115191	3/9/2019	1_BORN	0	3/9/2019	3/1/2019
115191	3/9/2019	PNEU.R	170	8/26/2019	3/1/2019
115191	3/9/2019	PNEU.R	173	8/29/2019	3/1/2019
115192	3/7/2019	1_BORN	0	3/7/2019	3/1/2019
131818	2/1/2019	PINKEYE	179	7/30/2019	2/1/2019
131820	2/2/2019	1_BORN	0	2/2/2019	2/1/2019

Re: Life events as a percentage of births

amitchandak — Wed, 18 Nov 2020 08:57:33 GMT

Anonymous , expected output is not clear to mean, Can you highlight and explain

Re: Life events as a percentage of births

Anonymous — Wed, 18 Nov 2020 15:13:54 GMT

// Careful: this measure honors all // filters you might put on the data, // so always make sure you understand // what it returns. T is the table // with the data you showed in your // question. Thanks. [Pneumonia %] = var __pneumonia = CALCULATE( DISTINCTCOUNT( T[ID] ), KEEPFILTERS( T[EVENT] = "pneu.r" ) ) var __born = DISTINCTCOUNT( T[ID] ) return DIVIDE( __pneumonia, __born )

Re: Life events as a percentage of births

Anonymous — Wed, 18 Nov 2020 16:56:27 GMT

Amitchandak
Thanks very much for your request for clarrification.

An Excel pivot table delivers these counts of health events with a grouping focusing on the birth month.

I then use the Excel Pivot specifications mentioned in the original post to determine the proportion affected (incidence rate) with the number born as the denominator.

Here's the specifications for the Excel pivot summarization:

The Pivot Table creates the incidence values using the fields of EVENT and BirthDate (summarized as a birth month BMON). In the Pivot Table the column is assigned to EVENT, and the rows to BMON. The summation of values is the COUNT of ID (the unique individual reference number), with the Show Values As option set to "% OF", Base field = EVENT and Base item set as 1_BORN. The resulting values are then displayed as percentages. Of all the individuals born in the month of Feb 2019, 3.5% encountered a pneumonia event is an example result.

This is an example of the desired output

The categories of 1_BORN, SOLD and DIED are constants between different datasets, while the health conditions can be variable in name.

I hope this better explains the output goal.

Thank you very much for considering this question!

KirkpM

Re: Life events as a percentage of births

Anonymous — Wed, 18 Nov 2020 17:06:35 GMT

daxer
Thanks very much for considering this question. I had another request to provide clarrification. Here's a further explanation.

An Excel pivot table delivers these counts of health events with a grouping focusing on the birth month.

I then use the Excel Pivot specifications mentioned in the original post (shown below) to determine the proportion affected (incidence rate) with the number born as the denominator.

Here's the specifications for the Excel pivot summarization:

This is an example of the target output from an Excel Pivot Table

The categories of 1_BORN, SOLD and DIED are constants between different datasets, while the health conditions can be variable in name.

I hope this better explains the output goal.

Thank you very much for considering this question!

KirkpM

Re: Life events as a percentage of births

wdx223_Daniel — Thu, 19 Nov 2020 03:29:19 GMT

Anonymous =DIVIDE(DISTINCTCOUNT(Table5[ID]),CALCULATE(DISTINCTCOUNT(Table5[ID]),Table5[EVENT]="1_BORN"),0)

Re: Life events as a percentage of births

Anonymous — Thu, 19 Nov 2020 21:46:58 GMT

wdx223_Daniel: Perfect! That is exactly what I was looking for and really solved a big stumbling block. Thank you! KirkpM