Showing all days within selection, even if no data / ValueLoop?

[#1]

I'm trying to create a table that shows all orders within the selected period (say, December 2009). We don't have orders on every day, but our customer would like to see every day (including weekends). I've got a Calendar table and an Orders table, which are linked with the date as a %DateKey. My graphs dimension is %DateKey, the expression is Sum(OrderCounter), where OrderCounter is 1 for each order. This creates a simple graph with an orders-per-day view.

I first tried to enable "Show All Values". This shows all days, including weekends without orders. Of course, if I select December 2009 it also shows all days before and after December 2009, which makes the graph unreadable. How can I limit the graph to only show December 2009 but still have "Show All Values" enabled? I've thought of generating a "fake" order for every day on which we don't have orders, but that would fail if the user selects a specific kind of order. The only option then would be to generate a fake order for every possible type of order we have (thousands!) for every day that those orders haven't been placed (almost every day in 4 years of data for most order types). We're talking millions of fake records just for the sake of a displaying all days in a graph.

I also came across the ValueLoop function, which appears to be perfect: throw in the minimum and maximum selected date (December 1st - December 31st) and it should generate 31 days for me. For some reason ValueLoop does not change when the selection changes. Does anyone know why, or what I should do to make ValueLoop use the current selection?

[lironbaram]

hi all you need to do

is in the properties of the chart, in the presention tab

uncheck the Supress Zero_values

attach a picture and example

[lironbaram]

hi

all you need to do is unchek the Suprress Zero Values in the presention tab in the chart propreties

here is Example

[#4]

Here's what I get with Suppress Zero-Values unchecked:

This is also with Suppress Missing disabled. The problem is that it's missing days (1, 6, 7, ..) while it should be showing those as 0 orders.

[#5]

Hi,

just create an additional formula with =1 and then makr it as invisible.

Best Regards HG

[#6]

This works so long as there's only 1 dimension. I've also got tables with multiple dimensions, if I add the same expression to those tables all dimensions are grouped into 1 big dimension. Is there anything I can do about that?

[#7]

Hi

try it with = 0 as Formula

HG

[#8]

The same thing happens, the result is a single line instead of multiple lines (for the second dimension).

[lironbaram]

hi

attach is an example file

can you produce an example table of the Data you load

hope its helps

[#10]

I've attached an example with data.

The top graph has no special settings, it ignores empty values, which means it sometimes skips a few days. To see this, select April 2007 and set the graph to show days (instead of Year-Month). If I remove the second dimension from this graph I get the same result as in your example, but I need the second dimension.

The bottom left graph uses the solution provided earlier (an extra expression with =1). It groups all OrderProcesses into 1 dimension, but apart from that works fine. Again, if I remove the second dimension this seems to work.

The bottom right graph is a little experiment of my own, it uses a Continuous graph starting at the first selected date and ending at the last selected date. This doesn't work perfectly, but it's closest to what I need. However, using the default selection it doesn't look right (2009-06 and 2009-12). Also, I can no longer view the graph by Year-Week, which I need as well. This is because Year-Week is really a string. I've changed it to a date using the WeekStart function, but then the Continuous graph shows integer values (40010 instead of "2009-42").