How to delete missing data and use the resulting database in a bar chart

Question

I want to construct a bar char for the following data, but I am having a problem with Missing[]:

countries = {"Afghanistan", "Bangladesh", "Bhutan", "India", 
   "Maldives", "Nepal", "Pakistan", "Sri Lanka", "Brunei Darussalam", 
   "Cambodia", "Indonesia", "Laos", "Malaysia", "Myanmar", 
   "Philippines", "Singapore", "Thailand", "Timor-Leste", "Vietnam", 
   "China", "Japan", "Mongolia", "North Korea", "South Korea", 
   "American Samoa", "Fiji", "French Polynesia", "Guam", 
   "Hong Kong SAR, China", "Kiribati", "Marshall Islands", 
   "Micronesia", "Nauru", "New Caledonia", "Northern Mariana Islands",
    "Palau", "Papua New Guinea", "Samoa", "Solomon Islands", "Tonga", 
   "Tuvalu", "Vanuatu", "Australia", "New Zealand"};
sAsia = {"Afghanistan", "Bangladesh", "Bhutan", "India", "Maldives", 
   "Nepal", "Pakistan", "Sri Lanka"};
seAsia = {"Brunei Darussalam", "Cambodia", "Indonesia", "Laos", 
   "Malaysia", "Myanmar", "Philippines", "Singapore", "Thailand", 
   "Timor-Leste", "Vietnam"};
eAsia = {"China", "Japan", "Mongolia", "North Korea", "South Korea"};
SIDS = {"American Samoa", "Fiji", "French Polynesia", "Guam", 
   "Hong Kong SAR, China", "Kiribati", "Marshall Islands", 
   "Micronesia", "Nauru", "New Caledonia", "Northern Mariana Islands",
    "Palau", "Papua New Guinea", "Samoa", "Solomon Islands", "Tonga", 
   "Tuvalu", "Vanuatu"};
Pacific = {"Australia", "New Zealand"};

data19 = {2.55, 8.05, 4.05, 5.25, 7.55, 7.15, 3.7, 2.8, 2, 7.25, 5.1, 
  5.5, 4.55, 6.6, 6.15, 2.4, 3.25, 0.35, 7.05, 6.25, 0.45, 6.2, 
  Missing[], 2.45, Missing[], 1.7, Missing[], Missing[], 0.8, 3.85, 
  4.8, 0.7, 3.35, Missing[], Missing[], 2, 2.8, 0.75, 2.55, 0.5, 4.6,
  3.1, 2.35, 2.9};
data20 = {-5, 3.8, -0.8, -8, -32.2, -1.9, -0.4, -3.6, 
  1.2, -3.5, -2.1, -0.4, -5.6, 3.2, -9.5, -5.4, -6.1, -6.8, 2.9, 
  2.3, -4.8, -5.3, Missing[], -1, Missing[], -19, Missing[], 
  Missing[], -6.1, -0.5, -3.3, -1.6, 0.7, Missing[], 
  Missing[], -10.3, -3.9, -3.2, -4.3, -0.5, 0.5, -9.2, -2.4, -3};
data = {data19, data20};
names = {"2019", "2020"};
ds = Dataset@Map[AssociationThread[names, #] &]@Transpose[data]

I like to use the database ds in constructing special charts in: https://mathematica.stackexchange.com/a/249707/60365

Initially, I gave the data as a list of lists and I had no problem. But now because of the Missing[] observations in my data, I realized that using the database ds is more advantagous but then I am stuck with how touse ds in the code given in the above link.

Answer 1

We can slightly modify the function arrowBar from the linked answer so that it returns {} for inputs with Missing elements.

ClearAll[arrowBarB]
arrowBarB[arrowwidth_: .5, boxwidth_: 1, arrowangle_: 120][
   colors_: ColorData[97, "ColorList"], dr_: 5] := 
  If[FreeQ[#3, _Missing], {colors[[#]], Opacity[.7], 
     ChartElementData["ArrowRectangle", {"ArrowheadAngle" -> arrowangle, 
        "ArrowWidth" -> arrowwidth}][{{-boxwidth, boxwidth}/2 + #2, #3}], 
     Opacity[1], EdgeForm[Gray], Disk[{#2, #3[[1]]}, Offset[dr]]}, {}] &;

We use preProcess and options from the linked answer as is and change the labels in legend:

ClearAll[preProcess]
preProcess[groupsizes_List, groupspacing_: 1] := 
  Join[Join @@ MapIndexed[
      Thread @ {#2[[1]], (#2[[1]] - 1) groupspacing + # - 1} &, 
      TakeList[Range[Length @ #], groupsizes]], List /@ #, 2] &;

options = Join[{ImageSize -> Large, Axes -> {True, False}, 
    AxesStyle -> Dashed, AspectRatio -> 1/2}, 
   FilterRules[Charting`ResolvePlotTheme["Business", Plot], Options[Graphics]]];

legend = Graphics[{arrowBarB[][{Gray}][1, -1/4, {1, -1}/2] /. 
     EdgeForm[_] -> EdgeForm[White],
    GrayLevel[.2], 
    Text[Style["2019", FontSize -> 11], Offset[{5, 0}, {1/8, 1/2}],  Left], 
    Text[Style["2020", FontSize -> 11], Offset[{5, 5}, {1/8, -1/2}], Left]}, 
    ImageSize -> 1 -> 40];

data3 = Normal @ Values[ds];

dt = preProcess[{8, 11, 5, 18, 2}, 3][data3];

groupnames = {"sAsia", "seAsia", "eAsia", "SIDS", "Pacific"};

groupcolors = ColorData[97] /@ Range[Length @ groupnames];

We color country labels red for countries with missing data:

missingpositions = Position[data3, _?(MatchQ[{_, _Missing} | {_Missing, _}])];

xticks = Thread[{dt[[All, 2]], Rotate[#, 90 Degree] & /@ 
     MapAt[Style[#, Red] &, countries, missingpositions]}];

We also add gridlines for each country to distinguish gaps due to missing elements from between-group gaps:

Legended[Graphics[arrowBarB[.1, .8, 120][groupcolors, 3] @@@ dt, 
  FrameTicks -> {{Automatic, Automatic}, {xticks,  Automatic}}, 
  ImageSize -> 800, 
  LabelStyle -> Medium,
  GridLines -> {dt[[All, 2]], Automatic}, 
  options], 
 Column[{SwatchLegend[groupcolors, groupnames], legend}]]