flipside, you have proposed something else still - that basically gives me two charts stacked on each other, no?
That also looks good - I'll include it in the portfolio of different visualizations I'll send to our CR-manager.
I now have the construct using two barcharts side_by_side looking quite good. I'll have to somehow make sure the dimension_axis doesn't change when I select another week which might, in combination with a month, have but two or three days. But I think that can be done, it's just a bit time_consuming as I guess it will involve some more line_objects with visibility_conditions.
One challenge I am now facing concerns the left of the two charts: How can I get the axis to use smaller intervals? 1.000 is a bit much, 500 would already be a lot better.
Ah, ok, I've got it.
Thanks a lot!
The second example is still only one chart so keeps the dimension axis consistent (note the dotted line to help the user determine which day the data values relate to). It allows for the user to see the figures more easily than my first example, but the first allows the user to compare/correlate the data values quicker.
I'm afraid you'll have to explain the thing about coloring the axes a bit more - I still can't seem to do it.
On the "axes" tab, in the top half, I have my two expressions of course, for both I have a colour_box towards the bottom left of the top frame. But well, my expressions (there are to "lorry#" actually, for early and late shift) have a colour assigned on the "formulas" tab and changing the colour in that box doesn't seem to have any effect.
What am I doing wrong here?`
Thanks a lot!
P.S.: Sorry, I didn't mention that at the beginning - I have 2 and 2 expressions to display, not just 1 and 1. That makes your first suggestion look a bit crowded, with two bars for every day and two lines. I'll try the second one.
Have you set a width, if set to 0pt it won't show. Also watch for the chosen expression setting back to the first one when you click Apply - always catches me out, that one.
Ah, ok, that works. But it only colours a line along the y-axis for me, not the numbers as it does in your example ...
Ah. Of course. I'll try that.
I just tried to get the scales a bit better populated - I cannot display the values in or on the segments, that gets too crowded to the point of illegibility. So the axis must suffice - but it was a bit meager, just having the values 0 and 50 in the lower half and 0 and 2000 in the upper half - but it did have a second 0_line in the middle.
<=> Now I have set min and max for all the axis and that works, but now I seem to have lost that second 0_axis which is unfortunate because I used it to display a red line to separate the two charting_areas ...
Ah, there it is again - it seems that the max of the axis must be greater than the largest figure there is to display, then that second 0_axis is displayed. Well, that might prove challenging going forward - this week, volume is extremely low, so the biggest number of lorries on one day is about 50 - in busier weeks, it might just as well be 100 and with min and max fixed for each scale, that will not be reflected in my chart ...
I do like that chart, it makes everything look neat and tidy.
I have just tried out a new variation of your chart because I guess our customer will not be that interested in the categorization of early and late shift, but jut want the total figures per day.
Now, in that chart, I can easily set the Design to "Stacked" and the columns in the lower half stack up neat.
<=> But I ´cannot stack lines and switching the two expressions in the upper half to columns make them stack onto the lower half and QlikView does no more keep any visible separation of the two halfs.
I guess an area_chart (with the complete area between the 0-line and the line coloured) is something completely different which I cannot combine into this, or can I?
To keep it simple, and assuming you know what scale your figures are, I would just divide one of the numbers by 10 (or whatever makes sense).
I do the same with my data all the time;
No. of Clients = 12
Total Spend = $10,000
Total Spend / 1000 = $10
(Just make sure you label appropriately - in this instance I would use the format "$0k" to get "$10k"
You can 'sort of' stack lines, you just need to create a total expression by summing the two line expressions. The colouring gets complicated though as the text on the datapoint has to be done in the expression under Text Color.