Take the 2-minute tour ×
Mathematica Stack Exchange is a question and answer site for users of Mathematica. It's 100% free, no registration required.

I am trying to create a chart similar to the view offered by TradingChart, but with custom data not relating to finance.

allData = {{"2014-03-31", 91.0657, 91.5555, Null, 79.4144, 90.0735, 86.9567, 87.1578, 79},
           {"2014-04-06", Null, 90.9839, 85.0053, 79.2423, 87.7565, 85.0126, Null, 108},
           {"2014-04-22", 90.9086, 90.9959, Null, 78.9365, 87.5949, 86.1197, Null, 93},
           {"2014-04-28", 92.3299, 92.068, Null, 79.7057, 90.187, 86.4088, Null, 98}}

I am using the first column as a date marker, the last column as a volume counter, and the remaining columns as data for the line graph. I should be clear that the size of these data sets are flexible. There may be more or fewer dates and there may be more or fewer columns for lines. I clean up the data a little then place it into a DateListPlot (on the actual plots I have included a legend and better markers, but that is not so important here):

temp = Drop[allData, None, {-1}];

lineData = {};
For[i = 2, i < Length @ temp[[1]], i++,
    AppendTo[lineData, Transpose[{temp[[;; , 1]], temp[[;; , i]]}]]
];

DateListPlot[lineData, 
    Joined -> True,
    PlotStyle -> Thick, 
    ImageSize -> Scaled[0.75], 
    PlotMarkers -> Automatic ,
    PlotRange -> {Automatic, {70, 100}}
]

Here is the output:

Graph

As you can see the data is sporadic. Some dates have no entries at all, some columns have Null values. It seems Mma is having some trouble with the sparse data. Ideally the lines would pass straight through to the next non-null entry, so some help with doing that would be appreciated.

But the main question I have is with regards to putting a bar chart underneath this that matches up with the dates and uses the last column of allData as the volume. It is not clear if there is any way to do this with Histogram or with BarChart and I am not so experienced with Mma to know if I have to do this with Graphics primitives.

Eventually I would like to add some interactivity, which is what drew me to the TradingChart:

CountBars

share|improve this question

1 Answer 1

up vote 3 down vote accepted

There is a very quick way to arrive at this. If you are new to Mathematica then maybe this is a starting point.

lineData = {allData[[All, {1, 2}]], allData[[All, {1, 3}]], 
  allData[[All, {1, 4}]], allData[[All, {1, 5}]], 
  allData[[All, {1, 6}]], allData[[All, {1, 7}]]}

volume = allData[[All, {1, -1}]]

p1 = DateListPlot[lineData, Joined -> True, PlotStyle -> Thick, 
  ImageSize -> Scaled[0.75], PlotMarkers -> Automatic, 
  PlotRange -> {Automatic, {70, 100}}];

p2 = DateListPlot[volume, AspectRatio -> 0.15, Joined -> False, 
  Filling -> 0, 
  FillingStyle -> 
   Directive[CapForm["Butt"], AbsoluteThickness[20], Blue], 
  PlotStyle -> None, ImageSize -> Scaled[0.75], PlotMarkers -> None, 
  PlotRange -> {Automatic, {0,Automatic}}];

Column[{p1, p2}]

enter image description here

There is a lot more you can do with this. For example to mimic the trading chart you may want to remove frame ticks on the bottom of p2 and add GridLines -> Automatic, GridLinesStyle -> Dotted to both plots. For regular use you will probably need to specify identical ImagePaddingin each plot to ensure that they line up. But this is a easy way to implement "on the run".

share|improve this answer
    
This is a great start, thank you. The only comment I have is that the For loop is necessary as I don't know a priori how many lines I will have. I could then add event handlers to the bars to handle the interactivity that I want, correct? –  Nick Jul 7 at 21:47
    
Yes but as part of learning Mathematica you should investigate doing this with Map rather than For ...or probably MapIndexed is the better option when you do not know how many lines. –  Mike Honeychurch Jul 7 at 21:56
    
I will look into MapIndexed as Map did not seem to be the best approach. Again, your advice has been quite helpful. –  Nick Jul 7 at 22:11

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.