Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

Cumulative Bar Chart

Hi,

I am looking to create a cumulative bar chart with a set of data. I've looked at other posts, but nothing seems to fit my circumstances.

I have a range of products and their installed dates:

Dimensions: Year, Month and Product Name

Expression: rangecount(above( TOTAL Count(EmployerProduct_id), 0, rowno(Total)))

The chart is set to 'Full Accumulation'

Essentially the issue I have is that where there has been no install taken place for any given day (which I converted into Year, Month), this does not show an accumulation of that record in that month (see .png attachment of screenshot).

I'd prefer not to use an inline table to populate the days where no install has taken place with a product record of '0' (which won't work anyway since I am counting the records, not summing) as new products are added to this periodically and it is a large enough data table already.

Can anyone help? Please let me know if you require further information.

Your help is appreciated!

Thanks,

M

5 Replies
Not applicable
Author

could you prepare small sample qv file with your data ?

Not applicable
Author

Hi Pari,

Please find attached a sample fo the data. The second graph is the cumulative chart, but as you can see where there have been no installations in certain year/months, it does not accumulate the product from the previous months.

In addition, once this is done I need to find a way to then take a product away from the chart at the point of it being decommissioned so that I can see a true reflection of the installed base for whatever month I am looking at. So if product A20 was installed in May 2007 but then decomissioned in May 2012 I would want to see it from May 2007-May 2012 in my chart but not beyond June 2012.

Thanks.

M

Not applicable
Author

Hi

sorry for late response...

1. You should create separate table for calendar with all the possible values CalcYM

2. ....still working on it....

Not applicable
Author

No problem, thanks!

I was afraid you would say that! So there seems to be no other way... Just to be clear I would need to create circa 500 records for every calendar day since 1970 and then maintain these when a new product comes along?

Thanks again.

M

Not applicable
Author

as i look at your data table, I suggerst to do crosstable load to make the table like:

Product_NameDateStatusDateType
A8724010In Service14/01/2007Calc
A8724010In Service14/01/2007CAT
A8724010In ServiceNullCommission
A8724010In ServiceNullDeliver
A8724010In ServiceNullInstall
A8724010In Service11/09/2006Ship
A8734986In Service16/01/2007Calc
A8734986In ServiceNullCAT
A8734986In ServiceNullCommission
A8734986In ServiceNullDeliver
A8734986In ServiceNullInstall
A8734986In Service16/01/2007Ship
A6514301Decommissioned17/01/2007Calc
A6514301DecommissionedNullCAT
A6514301DecommissionedNullCommission
A6514301Decommissioned17/01/2007Deliver
A6514301Decommissioned17/01/2007Install
A6514301DecommissionedNullShip
A6630855In Service17/01/2007Calc
A6630855In ServiceNullCAT
A6630855In ServiceNullCommission
A6630855In Service17/01/2007Deliver
A6630855In Service17/01/2007Install
A6630855In ServiceNullShip
A2497241In Service26/01/2007Calc
A2497241In ServiceNullCAT
A2497241In ServiceNullCommission
A2497241In Service08/05/2006Deliver
A2497241In Service26/01/2007Install
A2497241In ServiceNullShip

or even delete rows with Date=null

then try to create simple calendar (starting from 01-01-1970) with:

DateYM
01/01/197019701
02/01/197019701
03/01/197019701
04/01/197019701
05/01/197019701
06/01/197019701
07/01/197019701
08/01/197019701
09/01/197019701
10/01/197019701
11/01/197019701
12/01/197019701
13/01/197019701