Inga: You know, there's something I've been meaning to ask you. In the transference, the monster got part of your wonderful brain. But what did you ever get from him? 
Freddy: For the experiment to be a success, all of the body parts must be enlarged. Inga: In other words: his veins, his feet, his hands, his organs vould all have to be increased in size. Freddy: Exactly. Inga: He would have an enormous schwanzstucker. Freddy: That goes without saying. Inga: Voof. Igor: He's going to be very popular. by The FrankensTeam Few days ago a colleague of mine showed me a chart in Tableau which he put together to visualize total category value and underlying data (eg. subcategories). It was a nice and useful example to see the details behind summarized values. I replicated his example in Excel using column chart and error bars (those thin darkblue lines): It is a combination of column and xy scatter charts where I hide the scatter points (no marker) and add error bars with subcategory values. The base data table looks like: Because this is just a top view, I put the categories into columns with sorted subcategory values below. Using "dots start" and "dots end" cells you can position the error bars according to the width of the columns. All the calculations behind the chart is performed in named formulas. If you need more columns or rows, you can change the name "my_data". Update: Here you can read about another method to create similare chart. This post can help to use the template in your own workbook: How to copy charts based on named formulas to other workbook? 
Freddy: Damn your eyes. Igor: [to camera] Too late. by The FrankensTeam Well, when I had sent my solution to Igor about the combined column chart I made to visualize total number and underlying data, he sent me back an alternative one which... well, easier to set up, while visually almost the same. Here the "thin" columns are added as column chart  each row of the original data table is one data series. (You can easily do it: select the underlying data part as rectangular range, add column chart. Then go to the Select data menu and click Switch Row/Column button.) The trick is that: total columns are created using updown bars. You have to add two more data series to the chart: one for the category total values, and another with 0 values. Both are changed to line chart type. Now add updown bars to on of the line. Updown bars appear as columns between the two line data series. Unfortunately updown bar is always in the foreground of the chart, so it will cover the thin columns. To step over this problem, you can set the color of the bar with transparency, or use borders only. In our solution we chose gradient color. Width of the updown bar columns can be set under the settings of the line series of which the bars belong to. (Unfortunately this is not available in Excel 2007  you will need a row of VBA code to do it in this version.) You can download our file to see how it looks like. 
Igor: What is this? Freddy: Schwartzwalder Kirschtorte. The Monster: [offscreen] MMMMMMM! Freddy: Oh, do you like it? I'm not partial to desserts myself, but this is excellent. by The FrankensTeam This post is not about formulas. No charts. And it contains VBA. Sorry for that... it's a unique case, but I'm sure it can be useful for many readers... So, in the same time we publish it in Hungarian, I take the opportunity to make it available for those who do not speak my native language. :) In our example file and also on the below picture you can see a small pivot table showing monthly trend of Profit. We would like to give the possibility to the users to see the trend of another measure. Using slicer it's not necessary to open the Field List pane and look for the measure there  they can easily choose it by one click. If you choose more measures on the slicer, all will appear under Value fields in the pivot. Another advantage of this soultion is that you can limit the measures appearing on the slicer: for example you can offer only a subset of measures where the above pivot table with monthly trend can be interesting for analyze. Obviously slicers are for filtering a field and not for choosing field names... So to build up this solution, we will need a short and simple VBA code and an additional pivot table. Let's see how to do it step by step. First: the original data table, base of the above pivot. On the below picture you can see all the measure names highlighted by blue: Step 0: Create your pivot table. You will need the name of this particular pivot. You can find it under Pivottable Tools / Analyze on the left hand side. In the example the name is PivotTable1. You will also need the name of the worksheet where this pivot table sits. In the example it is named Pivot. Step 1: On a blank, new worksheet create a list of measure names (not necessary to include all the measures) and insert a pivot table on this list. Put measure into Row lables: Name the sheet Slicer. Step 2: Define Name choice with reference to the first cell under Row Labels. (Scope of the name should be workbook.) Step 3: Insert a slicer for measure, and move it to the worksheet where your original pivot table is. (Pivot sheet) Step 4: Open VBA Editor (Alt+F11). Find your workbook under Project explorer, and open the code module of the worksheet where the measure pivot table can be found  in the example it is the Slicer sheet. Step 5 Code for normal pivot tables: Follow this step if you use normal pivot table. In case of PowerPivot, go to Step 5 Code for PowerPivot! Put this VBA code into the code module of the Slicer worksheet. Private Sub Worksheet_PivotTableUpdate(ByVal Target As PivotTable) Dim ptMain As PivotTable Dim pfMeasure As PivotField Dim i As Long On Error GoTo Errorhandler Set ptMain = Worksheets("Pivot").PivotTables("PivotTable1") For Each pfMeasure In ptMain.DataFields pfMeasure.Orientation = xlHidden Next i = 0 Do While [choice].Offset(i, 0).Value <> "" ptMain.AddDataField ptMain.PivotFields([choice].Offset(i, 0).Value) i = i + 1 Loop Exit Sub Errorhandler: Debug.Print Now(), Err.Description End Sub If you would like to use the solution in your own file, you only have to change the worksheet name "Pivot", the pivot table name "PivotTable1" and the named range [choice]  pay attention this latest appears in two rows! Step 5 Code for PowerPivot: Put this VBA code into the code module of the Slicer worksheet. Private Sub Worksheet_PivotTableUpdate(ByVal Target As PivotTable) Dim ptMain As PivotTable Dim pfMeasure As CubeField Dim i As Long On Error GoTo Errorhandler Set ptMain = ThisWorkbook.Worksheets("Pivot").PivotTables("PivotTable1") For Each pfMeasure In ptMain.CubeFields If pfMeasure.Orientation = xlDataField Then pfMeasure.Orientation = xlHidden End If Next i = 0 Do While [choice].Offset(i, 0).Value <> "" ptMain.AddDataField ptMain.CubeFields("[Measures].[" & [choice].Offset(i, 0).Value & "]") i = i + 1 Loop Exit Sub Errorhandler: Debug.Print Now(), Err.Description End Sub Instead of PivotFields we use CubeFields, and a different string expression for AddDataField method. 
Igor: Oh, that'll be Inga. Herr Falkstein thought you might need a laboratory assistant temporarily. Freddy: Oh. Inga: Oh, hello, would you like to have a roll in the hay? It's fun. Roll, roll, roll in the hay. Roll, roll, roll in the hay. Roll, roll by The FrankensTeam There is always a debate if piedoughnut charts are useful or not. Well, technically this is an xy scatter chart. :) You can download the file and try it with your data. The animation is created based on the pivisualization animated chart: https://sites.google.com/site/e90e50fx/home/visualizemathsinexcel 
IntroductionOpen the dialog "Insert Hyperlink" and select the button "Place in This Document". A treeview allows you to choose a worksheet and type a cell reference but also displays all the names related directly to a range.However names that return a reference through a formula are not displayed and can not be used directly in hyperlink. How add hyperlink to a dynamic rangeAs in many other cases we can cheat the system security of Excel. In this case it is enough to change the name after it was added to the hyperlink.Here are the steps to follow:
A very interesting use is to combine this kind of hyperlink with validation list in a cell. In the below example we would like to illustrate how to create, in a few simple steps, a structure to navigate between worksheets in a workbook  of course without using VBA. How to create a structure to navigate between worksheets in a workbookThe whole thing will look as if you select the hyperlink itself from the validation list!First of all, you will need a list with the sheet names  write them somewhere on a sheet and assign a name “Sheet_list” to this range: When you have this name, create a validation with list on the cell you want to have the hyperlink. Important, that this cell must be exactly the same cell on all worksheets. In the example we use cell B1: Now you should create a name which will feed the hyperlink. First create a name only: “My_Sheet_Choice” with reference to the active cell  it will be changed later. Next step is to set up the hyperlink using the name created above. After rightclicking on the cell, choose Hyperlink, then click on “Place in This Document”. Here, under Defined Names you will find “My_Sheet_Choice”  that is what you need to select. And here comes the trick!  change the definition of “My_Sheet_Choice” to give back a reference to the sheet you selected from the list! In the Name Manager enter this formula to the Refers to field of “My_Sheet_Choice”: =INDIRECT(ADDRESS(1,2,,,INDIRECT(ADDRESS(1,2)))) (Please note, this formula will work only if you use B1 for the validation list. If you put the list to another cell, please change accordingly: the first parameter of ADDRESS is the row number of your cell, the second is the column number!) How does this formula work? The nested INDIRECT(ADDRESS(1,2)) reads the name you selected in B1 and gives it to the enclosing ADDRESS formula as 5th parameter = sheet name. ADDRESS combines the sheet name and the cell referenced by row and column numbers and results a syntactically correct reference as string (adds the apostrophes (‘) if needed and the exclamation mark (!) after the sheet name, so it will look like, for example: "'New York'!$B$1"). And here comes INDIRECT again, who is able to convert a string to a real range reference. When you enter the above formula into a cell, Excel evaluates immediately and gives back the value stored in the cell. But in our case we give it to the hyperlink as named formula. In this case no evaluation happens, the hyperlink will see the range reference and will navigate to the selected sheet. Now you have the hyperlinklist on one sheet. Copy it to all of your sheets into exactly the same range  the easiest way to do it is to select all the sheets (group) and paste on one sheet  it will paste to all the selected sheets. by FrankensTeam Send your comments to authors Add a comment to the discussion and read more tips on Linkedin 
Medical Student: But wasn't that the whole basis of your grandfather's work, sir, the reanimation of dead tissue? Freddy: My grandfather was a very sick man. by The FrankensTeam While building up my first PowerPiovot database, I realized a strange problem: tons of blank rows appeared in the final pivot table. Here I would like to share with you the details and how we can avoid it. Hungarian version of this post is available here. DatabaseThe problemIn the particular case it means thousands of rows mixed with rows with values. Obviously I did not want to give this "sick" report with "dead" codes to the users. InvestigationI realized the report contains data for these "dead" codes only in calculated columns, where I used a constant value in the DAX formula. (In the above picutre TMI% column is a calculated one.)And the sales data table: I loaded both tables to PowerPivot and joined with the Product ID field. You can see products number 7,8,9,10 are not represented in the data table. These products (and only these) belong to Product group x and y. I created the first calculated field for turnover growth (I used summarized measures): Growth % 1:= DIVIDE( [turnover 2015] , [Turnover 2014] , 0 ) Looks to be correct. But in our company the definition of Growth is different: we show only the value above 100%, so I have to subtract one: Using this new growth rate the pivot shows the groups without data too:Growth % 2:= DIVIDE( [turnover 2015] , [turnover 2014] , 1 )  1 Clearly the problem is the 1 constant value at the end of the formula. I tried with IFERROR too: Growth % 3:= IFERROR( [turnover 2015] / [turnover 2014] 1 , 0 ) The result is the same: pivot contains "dead" products. The only way I managed to eliminate the unnecessary rows is to change the calculation formula this way, using no constant value: Growth % 4:= IFERROR( ([turnover 2015][turnover 2014]) / [turnover 2014] , 0 ) What happened?Based on the join between the Product IDs, Excel builds up the calculation background table, containing ALL the Product IDs, but BLANK rows (without data) will not appear in the final pivot table.SolutionWell... I do not know what is the reason behind this strange way of using BLANK, it only makes problem for me. Fortunately the measures we use can be changed in a way without constant  this is the only solution I found for the problem. 
Freddy: Behind the bookcase. Hand me that robe, would you dear? You were right. It's coming from behind this wall. Where is it, where is it? Inga: What? Freddy: There's always a device. If I could just spot the triggering mechanism. Hello. It seems louder over here. Hand me that candle, will you? Put the candle back. Alright, I think i have it figured out now. by The FrankensTeam Have you ever wanted to connect the bubbles of a bubble chart with lines? For example to create kind of a timeline: how the position of a product changes month by month in a Growth/Profit matrix. You would like to see something like this: Bubbles are based on this sort of data table: Well, it’s a fact that bubble chart can not be combined with any other charts, so the usual “change series chart type” way will not work. Our outofthebox idea is to use trendline. We started to think about moving average trendline, because this is very simple to calculate based on the Period  as you can read on MS office site: “For example, if Period is set to 2, the average of the first two data points is used as the first point in the moving average trendline. The average of the second and third data points is used as the second point in the trendline, etc.” The idea is to create a dummy data series which determines a trendline which equals to the connecting line of the original data points. Only need to reverse the calculation logic of 2 period moving average: now the first data and the moving average (=the original data) is given, and need to calculate the second data. In other words, you will need a help column, where the average of the first two data is equal to the January Proft (2%), the average of the second and third data is equal to the February Profit (5%), etc. The formula behind this dummy series column now goes without saying. Obviously we need one data point more than the original series. This additional data point could be a fixed value or you could use the first data of your original table. Let’s see step by step.1. Build up the help columns.The veryfirst data (yellow cells) starts with a fixed 0.
2. Add a dummy series to your bubble chart using data from the help column: 3. Add a trendline to the dummy series:
4. Select the trendline and format it according to your needs. +1. Dynamic model It is easy to change the model to update automatically when you add new data below the data table. You only need to copy down the dummy series formulas to more rows, add names and use the names in the chart series instead of fix sheet ranges. In our file you will find the below names: x =booble_linea!$C$4:INDEX(booble_linea!$C$4:$C$102,COUNTA(booble_linea!$C$4:$C$102)) y =OFFSET(x,,1) size =OFFSET(x,,1) x_dummy =OFFSET(booble_linea!$G$3,,,ROWS(x)+1) y_dummy =OFFSET(x_dummy,,1) size_dummy =OFFSET(x_dummy,,2) You can download our example file for better understanding: In the next post we shared the story of this discovery and a bug we discovered about trendlines. Please read it! This post is available in Hungarian. Please share your thoughts with us! 
Wife: Harry, he was at it again. Harry: So what do you want me to do about it? Wife: Everyday. Harry: Let him, let him. Conductor: New York next, everbody out for New York. INT.  EUROPEAN TRAIN Wife: Hans, er dachte schon wieder. Hans: So? Wife: Aber jaden tag. Hans: Lass ihm, lass ihm. Conductor: Transylvania nachste. Jeder austeigen fur Transylvania. by The FrankensTeam From time to time the question of converting number to words raises on Excel help forums. If you google for the topic, you can easily find VBA solutions, but the problem could also be solved using Excel’s builtin formulas. Years ago in the Excel Hero LinkedIn group a challenge was opened by Daniel Ferry to create a formula for the English language conversion. The discussion and the “formula development process” was indeed interesting, and with the help of lots of great minds and strict eyes, we managed to create a relatively short formula. Recently, for another forum request, Igor put together the Italian language formula (plus UDF solution too), and just for curiosity, IngaKris created it for Hungarian language. Do you have formulas for other languages? French? German? Please slip your formula through slot in door, so we can publish it here! If not  please take a look at our files! Maybe you will be inspired to develop formulas for your language! Here are the files for download  please note there are some defined names (2 or 3) in the files. If you would like to use the formulas in your own workbook, please do not forget to copy the names! Ps: In Italian, for reverse conversion (undici > 11) there is a VBA solution using regular expressions created by Igor too. 
Freddy: Gone, gone, we've got to find him, do you understand? by The FrankensTeam It was just an experiment… First I tried with radar chart, but I realized it could not work. So finally we used stacked area chart combined with pie chart to make splot area perfect square. The calculation behind the chart is not complicated  we used the equation of ellipse to model the border of lighted and shaded parts.You can download our file to play with it. 
Harold: A visitor is all I ask for. A temporary companion just to help me pass a few short hours in my lonely life. Monster: Mmm. Harold: Thank you, Lord. Thank you. Monster: Mmm. by The FrankensTeam The ideaThe first inspiration to create a “circular cospmograph” came from Mike Bostock’s Uber rides by Neighborhood. Reading his article we learned the name of this visualization tool: chord diagram. We have seen lot of stimulating solutions on CIRCOS site.
We wanted to do something similar in Excel, although initially it seemd too difficult… but we are The FrankensTeam, we are able to resurrect the dead, so, for us, nothing is impossible. :) We started to build the first model using the technique developed for cosmographs. Only needed to modify the points of the lines using circles instead of vertical coordinate sets. After improving the formulas and clarifying the chart series, we managed to make the chart to be dynamic. We developed a great way to position the bands without crossing each other, so the final visual result is more clear.
Surfing on the net we found a better data set to create a good example for this new Excel chart. Gene Lu’s post is a critique about a Talent Traffic chart. Our version uses similar logic as his "Talent war bar graph breakdown" as stacked column chart: "Ratio is number of employees moving from Company A to Company B for every one employee going in the other direction.": You can see our version on the top picture. The color coding is the same as above, symbolizes where the talents came from. This way you can also see the ratio of employee left for example Yahoo (purple) because both the from and to side of the lines are visible on the same chart. In the Excle file mouse rollover technique is used to highlight the companies one by one, while the rest of the lines are lighter. Here you can see Microsoft highlighted:
You can download the file.
We created two simpler versions which are dynamic up to 10x10 data, so you can easily use the chart with your own data set.
On sheet MY_TABLE you have to write the category names to column A only, they will appear in the header row accordingly. Then fill in the table with your numbers, and... take a look at the chord diagram on sheet CHART!
You can download a rolloverbased dynamic version. Or you can choose a noVBA version too, where you can select the data series to be highlighted from a validation list. Update: we created a slicerversion of the chart also available under the download section. For the base chart we use as many data series as the number of rows in the data table  it is necessary to have different colors for the rows, but it makes the highlight technique a bit challenging. The trick is that we use the same data series for the highlight mechanism, but we change the data range behind the series. When you choose a category to be highlighted, only those lines are visible which goes to or come from the given category. On sheet SUPPORT we grouped the lineblock coordinates to support this highlight logic, so it was not difficult to build up the series using some OFFSET formulas.
You may also be interested:
Cosmograph dynamic version Edward Tufte's Slopegraphs in Excel Use slicers with interactive charts and dashboards Excel Gallery 
Inga: Oh, I'm so happy to meet you at last! by The FrankensTeam Our readers may know that the members of The FrankensTeam live in two different countries. Igor is Italian from Milano, IngaKris and DottorGábor are Hungarians from Budapest. We started to write together this blogsite 3 years ago, but have never met personally… until the last week. Milano was always on our bucketlist: Il Duomo di Milano, the fashion streets, The Scala and Leonardo’s Last Supper… must see! So we (Kris and Gábor) bought two tickets, booked an apartment and on 6th of March flying over the Alps we arrived to Milano in the early morning hours. The first impression about Italy was the Alps: the snowy peaks of the mountains were clearly visible lightened by the rising sun. The next day we met with Igor, his family and some friends and took a small trip in the mountains. We went up to a resthouse following a steam and ate great local specialities and drunk great wine and grappa there. All were very delicious! On the way down we followed another track where we had seen lots of nice wood statues: kobolds, turtle, drunken donkey, dragonfly, birds, love couple... labirints for the kids and a small museum too, with climbable wall. We spent a really pleasant day together! It was exciting to get to know each other personally, after this long virtual friendship! OK, but we are not a travelblog, so before our readers escape, let’s change topic and talk about a special Excel chart: The Alps mountain chart. Technically it’s a simple xy scatter chart based on coordinates of the peaks (more than 21k points!). We used 4 series with different color according to the height of summits. On sheet DATA you can set the limits of the series. Also you can filter for the countries*, series and blocks to check where they are on the chart. The summit we had seen from the resthouse is Corni di Canzo  you can find it in the downloadable file too. :) Here is a link to a dynamically generated panorama view from our viewpoint, from resthouse, where you can also see Corni di Canzo: http://goo.gl/ZQ9Nkj And the link to the panoramamaker tool for those who would like to play with it: http://www.udeuschle.selfhost.pro/panoramas/makepanoramas_en.htm For the chart data points we used this database: http://www.viewfinderpanoramas.org/ALPS.txt *not always indicate the country within whose borders the summit is situated. And… here we are! Ciao, The FrankensTeam 
Freddy: [...] Oh, I uh, you know I don't mean to embarrass you, but I'm rather a brilliant surgeon. Perhaps I can help you with that hump. Igor: What hump? by The FrankensTeam The inspiration for this post came from Visualizing MBTA data project by Mike Barry and Brian Card where we found a great interactive chart under chapter How People and Trains Affect Each Other. Here is a static screenshot about the chart: Quote from the original study: "The gray bands show the total number of entries into the all stations per minute over time for each day of the week." Each row represents one day, where the darker grey means more people. This graphic could be a good choice to visualize small and large data together, especially if you need to compare more data series. To replicate the chart in Excel is very easy. The chart type is area chart with more series  as many as the number of grey belts you want. You simply need to shift the original data by a chosen number called limit (according to the number of grey belts). For example imagine that your original data is scaled from 0 to 2000. Draw the original chart (the lightestgrey) and choose 500 as limit  so you will have 4 belts. Cut the chart to 4 pieces: below 500, from 500 to 1000, from 1000 to 1500 and above 1500. Now put the 4 pieces on top of each other. This is what you see in the orange frame below: In Excel you can realize it by shifting the original data down by 0, 500, 1000, 1500 and put the 4 series to the chart. Simply create a data table and minus these numbers from the original data. Create the area chart with 4 series and set the axis limit accordingly: minimum 0 maximum 500. Note: This way the slope of the areaparts will be correct. If we changed the data above 500 to 500 and below 0 to 0, the slope would not be correct because the lines would start from incorrect point on the horizontal axis. We created an example using traffic data about the famous Hungarian “Line 1” which is the second oldest underground railway in the world (opened in 1896) and listed as UNESCO World Heritage. Data (col. C) and chart series (EH) of Station 8 (Bajza utca  Hősök tere): and chart of Station 8: (enlargement of the above orange frame) As you can see, this kind of chart clearly displays the details for the whole range of values. The color emphasized 'wrapping' allows the easy visual identification of the higher value sections. The final chart showing all the stations together looks like this: We simply copied the charts to one sheet, and put on oversized chart with gridlines, legend and horizontal scale to the background. Now you are wondering why Croissant is in the title? Gábor named the chart because its layers are similar to this bakery product (in Hungarian we name it “kifli”) Kris published a tutorial in Hungarian too. You may also be interested in: 
Freddy: Dressed up like a million dollar trouper Trying mighty hard to look like Gary Cooper Monster: SuperDooper [...] by The FrankensTeam Charlie asked us to create a special version of the matrix bubble chart where the bubbles are colored red, green or orange according to value limits: red if data <10 green if data >=100 the rest is orange. In a chart colors could be assigned to data series, so we will need separated series for the colors. First we will define 3 new names following the color rules: arr_r=rng*(rng<10) arr_g=rng*(rng>=100) arr_o=rngarr_garr_r (The last letter of the name stands for the color.) Now we can add the new data series to the chart,  for example here is the red series: X and Y values are the same, the new names go to bubble size. (In case of the green series, it is enough to modify the already existing series with the new bubble size.) The new series must be assigned to primary axis and scale bubble size to 50, and certainly choose the colors according to your needs. Also needed to add labels to the orange series. You can see these steps in the original post. The result: We created a general version where you can easily change the limits and rules for the colors: Here are the names used in this version: arr_r=rng*((rng<'1'!$C$3)+('1'!$B$3="<=")*(rng='1'!$C$3)) arr_g=rng*((rng>'1'!$C$5)+('1'!$B$5=">=")*(rng='1'!$C$5)) arr_o=rngarr_garr_r Download our example file. If you plan to use this chart in your workbook, here you will find how to copy it from our example file to your own file. 
Freddy: If you're blue and you don't know Where to go to, why don't you go Where fashion sits? Monster:
Puttin' on the Ritz [...]
IntroductionIn the past we alredy have tried to represent all the data of a table in a single chart (check the Stacking Cubes Charts with Excel article).
The purpose was to give an overview, a total picture without focusing on anything in particular. This is just the beginning then you can change the formatting and customize the chart, with color and highlighting what interests you most or use in combination with others to create a dashboard.
Download the filesStep to step the creation of the Excel chartWe start by defining a dynamic range, for example, in this way:
In the picture below you can see what we want to realize. To achieve this we must add 3 series, one is to show the bubbles, the other two are used to show the column and row labels. The chart could easily be replicated based on your own data. In a nutshell: For the bubble positions, we use fixed x and y coordinates, the bubble size is the original data.
A little trick with the label formatting helps us hiding the lower values (<50 in this case) which, as is shown above, would lead to confusion. On the below pictures you can see the defined names and the three series used on the chart, so you can easily follow the construction steps.
matrix_data serie: arr_h=rng arr_x=(COLUMN(rng)MIN(COLUMN(rng))+1)*(rng=rng) arr_y=(ROW(rng)MIN(ROW(rng))+1)*(rng=rng) axis_s serie: asse_x_val_h=asse_x_val_y+1 asse_x_val_x=OFFSET(rng,1,,1) asse_x_val_y=COLUMN(rng)*0 You need to notice that in the serie "axis_x", x values are the column headers of the table (text value, if you look the file will see that the values in the cells are '2008, '2009 ... '2013) and that both series "matrix data" and "axis_x" are backed on the primary axis. Finally the bubble size is the original data, but to avoid overlapping circles, set the Scale of the bubble size to 50, so the diameter of the largest circle will always be 1.
To build a yaxis that displays the row headers define the following names and the "axis_y" serie: asse_y_val_h=ROW(rng)^0 asse_y_val_x=OFFSET(rng,,1,,1) asse_y_val_y=ROWS(rng)INDEX(arr_y,,1)+1 But be careful! in this case we associate the series to the secondary axis ...
The last step is to set the maximum value of the axis to a very large number:
If you add new data (new columns or new rows), the chart will resize . In the picture below we have added the row with the totals of each year.
Dynamic version based on the activecellFor a dynamic version based on the activecell use this code:
'in the Worksheet class module Private Sub Worksheet_SelectionChange(ByVal Target As Range) Static b As Boolean If Not Application.Intersect(Target, Range("Rng")) Is Nothing Then ActiveSheet.Calculate b = True Else If b Then ActiveSheet.Calculate b = False End If End If End Sub We have to add a new serie to represent the current value (a single bubble with black outline).
Use these new names: check_=AND(CELL("row")=MODE(ROW(rng),CELL("row")),CELL("col")=MODE(COLUMN(rng),CELL("col"))) active_h=IF(ISERROR(check_),0,INDEX(rng,CELL("row")MIN(ROW(rng))+1,CELL("col")MIN(COLUMN(rng))+1)) active_x=IF(ISERROR(check_),1,CELL("col")MIN(COLUMN(rng))+1) active_y=IF(ISERROR(check_),1,CELL("row")MIN(ROW(rng))+1) In the picture the result:
The value corresponding to the active cell is highlighted and a label describes the detail of the data. When the active cell is outside of the table, it will disappear. If you plan to use this chart in your workbook, here you will find how to copy it from our example file to your own file. If you like this chart, you should take a look at the version we created for a user with conditionally formatted bubbles. 
Fraü Blucher: Would the doctor care for a brandy before retiring? Freddy: No, thank you. Fraü Blucher: Some warm milk, perhaps? Freddy: No, thank you very much. No thanks. Fraü Blucher: Ovaltine? Freddy: Nothing, thank you. I'm a little tired. Fraü Blucher: Then I will say good night. Freddy: Good night. by The FrankensTeam We decided to publish a short post when we receive the 100th comment in our Gallery Visitor’s Book. Actually when I found time to write, we already have 102 comments. :) Freddy: Reputation. Reputation. Looking back to the history of the Gallery, we had 40k visitors  well, we are not Louvre, but we are happy with the results. All the comments  and the lot others we received in email  made our days and proved that what we do is interesting, useful and thoughtprovoking for others. Unfortunately in the past months we did not have time to work on new charts and new posts… Inga: Dr. Fronkonsteen, are you all right? Yes, we are all right, and we still have some ideas. So let’s hope soon we will return to the Excellife, because we have some great projects in the “Under construction” folder. :) At the end of this short note, let me quote from a nice mail  about our team: “And I say to myself, <<Of course they are. They are from Budapest & Milan. How could they do any less?>>" Villager: He's a Frankenstein, and they're all the same. It's in their blood, they can't help it. All those scientists, they're all alike. [...] 
Igor: What hump? [...] The IdeaA few days ago I tumbled on the Washington Post site about President Obama's 2014 budget proposal. I found the visualization really interesting  the squares represent the most important parts of the budget, and I can see the details of each part using the same square structure. Amazing and dynamic  I immediately wanted to create it in Excel and i thought «It  could  work!» :) The first question is: how the squares are placed? The base idea is that the squares ordered from largest to smallest are placed in the corner closest to the origin of the axeses. The methodology is described here  it is an interesting problem solved with an algorithm. I am sure that this view is suitable to illustrate a few data from 23 to a maximum of 1012 (In my final file according to the original article shows up to 17 values). If the maximum number of squares is known and the squares are ordered from largest to smallest it is possible to simplify the positioning problem by assigning fixed positions to the squares. So I assigned the positions to all squares in addition to the 8 you can see in the above picture. Square chart with Excel chartBased on the positions, I had to compute the coordinates of the squares. It is easy using the length of the edges, and the edges are the square root of the data value according to the original article where they use the area af the square to represent the data. As I thought about how to draw squares using coordinates, an old chart made with error bars come into my mind. This way it is rather easy to create the chart  this is the base idea: So only the rightbottom and topleft coordinates are needed. Rightbottom coordinates are determined by the fixed position and can easily be calculated by adding the edges of the appropriate squares. I name these coordinates as (x, y)  you can see marked with red circles on the below picture. The topleft coordinates (x1, y1) can also be calculated easily: x1 = x  edge y1 = y + edge Marked with green circles on the below picture. To draw the sides of the squares, I will use the error bars with the value of the edge  only the sign and the direction is important:  Starting from the rightbottom corner we need the verticalpositive and the horizontalnegative bars.  Starting from the lefttop corner we need the verticalnegative and the horizontalpositive bars. As you see the whole structure is very simple, the only operation used is SQRT to calculate the edges and adding/substracting them. The most important part of the drawing is done by the Error bars. (Kris was amazed by this tiny thing  Error bar was totally new for her. Later on she was inspired to create some funny charts using it... maybe we will publish soon. :)) To make the visual result more readable, we can create a third series to display the labels in the center of the square using these coordinates: x2 = x  edge / 2 y2 = y + edge / 2 Unfortunately to color the squares a workaround is needed. I filled the them with random lines using xy scatter chart. After making some fun with it.... ummm.... mainly Kris :) I threw out the idea. I really missed the effect of changing the selection dynamically and show the details as moving the pointer on the graph. It is pretty easy to do with js and forms but not in Excel. The solution I wanted to use is the mouse rollover technique. It requires the cells below the chart area to be sized according to the elements of the chart, so the chart must be static. If I fix the size of the squares I will lose the possibility of changing the data. Dead end. :( Square Chart with conditional formattingSo I tried another way to visualize filled squares: Conditional formatting of cells. The logic is very similar, but instead of drawing the squares on a chart I colored them directly into the spreadsheet using conditional formatting on cells resized to small squares.I use 100x100 small cells to map the structure  in a support sheet I compute indexes  so where a square should appear, I have numbers >0, and I have 0s for the white space around and between the squares. I place the “fake chart” on a new sheet because this way it is easy to use the mouse rollover technique to make the whole construct dynamic. As you move the mouse above the squares, a dynamic text shows the name and amount of the actual square or lists all the budget parts if you position the mouse outside the squares. On the righthand side a smaller chart appears illustrating the breaking down of spendings by category. To have a full picture, I created the Revenue part too, so you can choose between spendings and revenue from a list. To obtain all these dynamism, in the file we use only 6 (six!) lines of code: two VBA functions used within the HYPERLINK formula. I think and I am increasingly convinced that the rollover technique discovered by Jordan has revolutionized the way in which the users interact with Excel dashboards making them looking and working similar as graphics made with js. All the rest of the file is made with formulas and tools available to the average user. To be honest I have not spent a lot of time optimizing all the gears ... despite this, it turns like a good Swiss watch :) I have also prepared a simplified version to make it easier to adapt to your data. In this file we used color scale in the conditional formatting and added a dynamic label based on the colors. Dowload the file: President_Obama_s_2014_budget_proposal.xlsm Square_chart_with_cf.xlsm Treemap with conditional formattingAnother way (a “usual” way) to visualize this kind of data is the wellknown treemap chart. You can read about it here and you can see my old files here.For this project I built up the treemap using conditional formatting and the dynamic rollover technique: I made the legend dynamic too, so when you move your mouse over the color labels, you can see the square highlighted on the “chart”. The files are quite big and contain resourceconsuming formulas  to enjoy it you need a pc say "modern". :) I created the files with Excel 2010 and tested in 2007. To use it with Excel 2003 it will definitely need some modifications. Dowload the file: TreeMap_with_cf.xlsm 
by The FrankensTeam As an answer to a frequently asked question, we would like to share a tutorial about a simple method to copy a chart based on named formulas to your own file and use it according to your needs. Our charts in E90E50charts gallery are usually built on named formulas. This way we do not use sheet cells to store the calculated background data, so the sheet is clear: the reader only needs to change the base data and labels in the downloadable example file and the chart is automatically updated. Unfortunately the example file is not suitable for those who would like to use the chart in an existing workbook. In this case the simple copy and paste the chart method does not work since the named formulas will not be created in the other file. Because of the complexity manual creation and relinking the names is usually irksome and time consuming. So how should you copy the chart? The trick is to copy the whole sheet and replace the link under the Data menu. Here is how to do it stepbystep using the matrix bubble chart as example. 1. Open the downloaded example file.In the Name Manager (Ctrl+F3) you will see there are some named formulas of which rng is the dynamic named range of the data table.2. Open your file where you would like to use the chart.In this tutorial the file we use is My_workbook.xlsx. (It is important you have to use an existing file, alternatively you have to save the file before going to step 4.) 3. Move the sheet with chart to your workbook.In the example file rightclick on the sheet tab and choose Move or Copy In To book select your workbook and check the Create a copy box. Now the sheet is in your file, and the named formulas are created accordingly  you can check it in the Name Manager. The problem is that the chart series are linked to the original file, as you can see in the chart series formula: 4. Change the data connection link source to your fileUnder Data menu / Connections group / Edit Links click on Change Source and choose your own file from the Change Source dialogue box. (in the example it is My_workbook.xlsx)You will see as if nothing happened on the Edit links pane, but now the chart series references are correct, showing the name of your file: After you Save, Close and reopen your file, the link to the original file will disappear from the Edit links pane too. Now you can use the chart in your workbook without any problem, and you can change the reference of the data table. 
Freddy: What a filthy job. Igor: Could be worse. Freddy: How? Igor: Could be raining. by The FrankensTeam Introduction[...] Back in 2004, Edward Tufte defined and developed the concept of a “sparkline”. Odds are good that — if you’re reading this — you’re familiar with them and how popular they’ve become. This is the introduction of Edward Tufte’s “Slopegraphs” by Charlie Park, if you have not already done so you should definitely read it. One of the strengths of Tufte's graph (perhaps the most important) is that we can immediately read the ranking of both the data. Another of the strengths is that with this chart we provide a lot of information using very little ink and colors  in a really clear structure. Both of these points are problematic if you want to replicate Tufte's graph with a dynamic chart in Excel. Peltier Tech in his Slope Graphs tutorial proposed a chart using line chart type, but both of the features are missing from his solution. In this post we would like to share with you how we created slopegraphs in Excel solving the above mentioned two challenges:
On the below picture you can see what we would like
to create: Here you can download the tutorial file and the general solution with mouse roller. If you plan to use the tutorial chart in your workbook, here you will find how to copy it from our file to your own file. First we will create the chart in a static way for better understanding, then we will make the whole chart dynamic using dynamic named ranges. You have two data series for the two columns, and an xaxis numbered from 1 to 10 instead of the text data of column A: Add data labelsDo it for both of the series. This is what you see when the above settings are done: This is what you have after setting the axis maximums to 1000000 and set the secondary axis values in reverse order. The labels series appear as a vertical line, and the series linked to the secondary axis is placed on the righthand side, according to the 0 value of the axis: We're almost done. Now we add the data series using named formulas, and make some cosmetical changes. Add the slopes data seriesBecause both of the two possible axes are used for the labels, and has a maximum value 1000000, so we have no other option, we must rescale the data to be compatible with the primary axis. We will calculate a step value based on the number of columns in the data table and modify the x values with it. We will need the following names: rng= $B$2 : $C$11 step=1000000/(COLUMNS(rng)+1) x_1=(IF(MOD(ROW(rng),2),COLUMN(rng)MIN(COLUMN(rng)),MAX(COLUMN(rng))COLUMN(rng))+1)*step y_1=IF(MOD(ROW(rng),2),rng,MMULT(rng,(TRANSPOSE(COLUMN(rng))=MAX(COLUMN(rng))COLUMN(rng) +MIN(COLUMN(rng))))) Add new data series to the chart using these names: Here is the result: delete the axes and the grid remove the colored lines from the labels series add labels (using x values) to the date serie... ...now change the color of the series and reduce the thickness of the line Hide the vertical linesTo do this we use a new series, in the same color as the background of the plot area.We have to add two new names: x_w=(CHOOSE({1,1,2,2},0,MAX(COLUMN(rng))MIN(COLUMN(rng)))+1)*step y_w=CHOOSE({1,2,2,1},MIN(rng),MAX(rng)) Add a new series using these names: and VOILA'! Make the chart dynamicThe graph is ready to be done dynamic, what remains to be done is to change the name rng to a dynamic named range which will always contain the full data table (rows and columns): rng=OFFSET(Data!$B$2,,,COUNTA(OFFSET(Data!$B$2,,,100)),COUNTA(OFFSET(Data!$B$2,,,,100))) Based on rng create the following names for the labels series: label_x=OFFSET(rng,,1,,1) label_y=INDEX(rng,,1) label_y_right=INDEX(rng,,COLUMNS(rng)) Change the range references in x and y values of the labels series in the Edit Series menu: Replace range A2:A11 with label_x (in x values of both series) Replace range B2:B11 with label_y Replace range C2:C11 with label_y_right In the next examples we use the mouse rollover technique to highlight particular data on the chart. As you move the mouse over the data table, the label and the line will be bold. The TableGraphicThe below chart image was made after selecting the value 39.0 of Britain in the year 1979 in the data table:(Tufte, Edward. The Visual Display of Quantitative Information. Cheshire, Connecticut: Graphics Press; 1983; p. 158) Download Excel FileCancer survival ratesThis one is from Tufte himself. It shows cancer survival rates over 5, 10, 15, and 20year periods using 4 data columns.From: http://charliepark.org/slopegraphs/
Actually, the chart is a refinement of a Tufte original (2002), done (again) by Kindly Contributor Dave Nash (2003, 2006). This version uses two colors to separate the two groups of countries based on the change in infant mortality rank positions. The background data is separated into two groups.
More explanationsIn the background we use some interesting tricks about which we will tell more details in this post:
2D matrix as series for slopegraphsWe built up 2D matrixes for the x and y values of the data points. You need to know that when using matrixes on xy chart, the data points are connected with lines from left to right:As you can see there are crosslines on the chart when the last data of a row is connected with the first data of the next row. It makes the chart unreadable. To eliminate the crosslines, it is needed to read the data this way, as a serpentin: So the first row is readed from left to right, the second row from right to left, and so on. Because Excel is not able to do it, we have to change the structure of the data by reversing every second row in the matrix. By doing it, Excel’s “reading” logic will result what we wanted. Using this modified matrix on the chart, all connecting lines will appear on the sides: Now we have to build up this matrix using excel formulas. Formula challengeThe key element of the chart is the matrix with reordered data as you can see above. Creating this matrix is an interesting formula challenge. On the below picture the blue table is the original data of which we would like to reverse every second row:We will create a help matrix first, then try to build a formula where we will use the first row of the original matrix, then the second from the help matrix and so on. The help matrix is relatively easy to create if you are familiar with the rules of matrixmultiplication (in Excel: MMULT). If we multiply the original matrix with an antidiagonal 01 matrix, the result will be the helpmatrix, all the rows reversed: Using the name rng for the original data table, it is not difficult to create the antidiagonal matrix: (TRANSPOSE(COLUMN(rng))=MAX(COLUMN(
So the formula creating the help matrix: =MMULT(
And now, we only need an IF formula: if the row number of the original data table is odd ( MOD(ROW( rng ),2)=1 ) we use the original matrix, if not, we use the help matrix:=IF(MOD(ROW(
You can realize that the result of this formula will be different if the data starts in even row or odd. In our case, this result will be used on the chart as y value, and we will build an xvalue matrix in a very similar way, so on the chart this difference will not appear. If you would like to fix the structure of the result, you will have to normalize the row numbers starting from 0, so in the MOD function you will need this change: MOD(ROW( rng )MIN(ROW(rng )),2)In the example file the above IF formula used as part of the named formula y_1. 
Dr. Frederick Frankenstein: For the experiment to be a success, all of the body parts must be enlarged. Inga: In other vords: his veins, his feet, his hands, his organs vould all have to be increased in size. Dr. Frederick Frankenstein: Exactly. Inga: He vould have an enormous schwanzstucker. by The FrankensTeam We would like to show some tricks in Excel charts how to focus on a small change in a large population. The target is visualize the trend without overwhelming it by the total value. The problemHow to visualize the growth of the proportion of women CEOs in Fortune 500 companies? This is a typical problem of large numbers: the population consists of 500 companies, while the number of women changes from 0 to 20. The change is significant, so there is a need to visualize it but also highlight that it is a only a small part of the total. The base idea is to split the total data to more rows to make the maximum bound of the chart axis lower  this way the important small values are clearly visible. The remainder range above the small value appears only in the last year.Igor posted an example based on this idea in this discussion on Alberto Cairo’s site. Here is the solution made in Excel: On the chart you can see the growth from year 1996 to 2013 but also have an impression about the total 500 if you take a look at the grey crowd. Excel chart solutionsThis chart could easily be done in Excel using a stacked bar chart type where the bars are filled with pictures.We created two versions, you can download both, and below you can read a stepbystep guide about the static version. Download the static chart Download the dynamic chart Static version step by stepLet’s start the explanation with this simplified example:The total data is 500: it will be splitted by 100, so we will have 5 rows and 100 elements in each at the bottom of the chart. This represents the total. The data of the last year is combined with the first 100 element to make the comparison easier. Here is the data table and the series data used on the chart: If you take a look at the numbers, you will understand how the red and grey series data was calculated. Using this data table you can easily create the chart:
Now you can see something similar: The next step is to fill the lines with pictures. We used these two pictures: (In Excel 2013 we realized that the chart displaying is extremely slow if you use large sized pictures, so we suggest to use relatively small ones. The above size worked well for us.)
Do the same steps with the second series and the grey man picture. You will have to change the Gap Width of the series to a relatively low number to have wider bars and betterlooking pictures. (You can observe that using this fill technique, Excel is able to illustrate decimals by cutting the last small picture within the bar.) Now the static version is ready to use, but you can download our version too. Dynamic versionTo change the chart to be dynamic you will only need to implement some named formulas and change the chart series to point the named formulas instead of sheet ranges.Using this version you can add new data and can change the total and split by values. (In our file all the yellow cells are input cells.) You may need to change the horizontal axis maximum bound according to the Split by data. On top of that, we added an extra feature: the possibility to create gridlines only for unique values. It looks like: Now, instead of going into details about the dynamic technique, you can download the file and learn it. If you plan to use this chart in your workbook, here you will find how to copy it from our example file to your own file. 
Freddy: Aren't there any lights in this place? Igor: Two nasty looking switches over here, but I'm not going to be the first. by The FrankensTeam IngaKris found a really eyecatching infographic by another Hungarian Krisztina: Krisztina Szűcs, about film production profitability, and she immediately wanted to do it in Excel. http://szucskrisztina.hu/images/holly.png Here is the final result with some Mel Brooks film (including Young Frankenstein of course!): So no, it seems to be, but it's not an area chart! On the left scale you can see the score of the film (Rotten Tomatoes). The light effect from this "projector" shapes the profit on the right side: the lower point is the budget, the upper is the gross income. Each of the triangles formed by 50 lines going
across the plot area. The base of technique is explained under an
earlier post about slopegraphs. All the formulawork is done within named formulas. On the chart we use separated data series for each of the films, because the transparencycummulation effect does not work if the crossing lines belong to the same series. Another advantage is the possibility of using different color for the series, and also, the formulas are easier to understand. Scaling The tricky thing is in the scaling: on the left side of the chart you can see the Rotten tomatoes score of the film: the value is between 0 and 100. On the right scale two values belong to each film: the bottom value is the budget, the top value is the worldwide gross income. The difference is profit and it appears as the side of the triangle... or the lightspot on the screen. Score and profit is measured on different scale, so to draw the chart we will need to rescale the scores according to the profit  that's the reason why we put the scale max values on the sheet and use these (as named cells) in the formulas. The axes of the chart must have these values as maximum. Named formulas The named formulas behind the chart are simple. The number of lines used to draw one triangle is set as a constant: lines = 50 Also we created names score_max and profit_max for the maxvalue cells.A range of 50+1 rows and 2 columns... lines_range = OFFSET(Film!$A$1,,,lines+1,2) ... will help us to build up an array for the xcoordinate values. It must look like: 1; 2 2; 1 1; 2 2; 1 etc. for the back and forth lines between the left and right side of the chart. (more about it here) The named formula to create this two dimension array is: x_0 = =IF(MOD(ROW(lines_range),2),{1\2},{2\1}) Now, for the 1 values (left side) we have to assign the rescaled score value of the film. For 2 (right side) we have to assign the budget and profit interval, split to 50 steps. The y value formula of the first film looks like: y_1 = IF( x_0=1, Film!$C$6/score_max*profit_max, Film!$D$6+(ROW(lines_range)1)*(Film!$E$6Film!$D$6)/lines ) Label trick We added one series for the white dots (movie projectors) plotted on secondary axis. We use the original score values and scaled the axis accordingly. The trick is that we use the labeltext column as x values of this series. Excel will interpret the text as value, starting from 1, so the dots will not appear in correct position. The solution will be to set the scale maximum of the secondary x axis to a very large number. And the labels could be added using the x values of the series! You can find a detailed, stepbystep description of this label trick in our post about slopegraphs. If you need more series form more films If you would like to add a new series (new film) to the chart, you will have to build a name for the y values. The easiest way is to copy the formula of an existing y name, and change the cell references. Than you can add the new series using x_0 and the new y name. For the labels, simply expand the label and the score value ranges. You can download our chart to learn it. If you plan to use this chart in your workbook Here you will find how to copy it from our example file to your own file. You may also like and learn: Stacked column flow chart with Excel Edward Tufte's Slopegraphs in Excel How to use rectangular range as chart data 
Dr. Frederick Frankenstein: Well it seems as if our mysterious violinist has disa... [sees something] Dr. Frederick Frankenstein: puh. Inga: Disa what? Igor: ppeared. Dr. Frederick Frankenstein: Shh. by The FrankensTeam The problemSometimes we would like to use rectangular (2dimension, 2D) range as chart data especially when the input data is not / could not be placed in one row or one column. Those charts which designed to plot only one value data (column, line, area chart types, etc) always use the horizontal (x) axis as a category axis, where data is distributed evenly. Only the vertical axis is a value axis and represents values. In this case Excel will plot the rectangular range by series. You could choose if you would like to use the rows or columns as series, so you will have as many series as many rows (or columns) you have in the data table. The visual result depends on the chosen chart type. The most important common point is that all the series will be colored different. So for these charts 2D range is not a problem, however you may want the data to be plotted as one series. It could be useful when you would like to use only one color on the chart, without setting it manually for all the series. In this article we will show you an easy trick how to do it, then we will give some illustrated examples to show you the usefulness of this trick. The real problem of 2D ranges come out in case of charts with two value axes (xy scatter, bubble). Excel Help tells us about xy scatter charts: “To arrange data on a worksheet for a scatter chart, you should place the x values in one row or column, and then enter the corresponding y values in the adjacent rows or columns.” http://office.microsoft.com/enus/excelhelp/availablecharttypesHA001233737.aspx#BMscattercharts That means we have to find a workaround if the data looks like: If we try to create a blank xy chart and add a new data series, Excel will allow to select x values as a rectangular sheet range with more rows and columns, but this is just an illusion. (Without selecting any data: Insert / Chart / Scatter, then right click on the chart, choose Select data and push Add) X axis could be treated as category axis in xy charts too. If Excel is not able to recognize the x values as plottable values, it will use your data as concatenated text: you can see on the picture 2 3 4 is separated by space character. This text string will be distributed evenly along the axis  similarly to a line chart  so the result will not be correct! (Fortunately) when you try to do the same with y values  you will get an error message: The sheet range for y values could definitely be only one row or one column. The key of the solution is that charts work with array of numbers, but Excel does not automatically convert 2D ranges to arrays. We have to create named formulas to do this conversion “manually”. So go to the name manager (Ctrl+F3) and create two names: Of course you can use different references for x and y values: you can use a single row for x values and a single column for y values  the important thing is the number of elements. Also you can combine named formula and range reference similarly.x_array=Sheet1!$B$3:$D$7*1 y_array=Sheet1!$B$11:$D$15*1 Important that it is not enough to use names for the data ranges: an operation must be done to force Excel converting the data from range to array of numbers, that is why we use multiplication by 1. After defining the two names, we can use it as series data. Simply go to the Edit series menu, and write the names to the x and y values fields as you see below and the chart will work as we wanted: If you would like to connect the data point with line, you need to know that Excel will draw one continuous line row by row, so the last data of a row will be connected with the first data of the next row: You will have exactly the same chart if the data comes from one row: In general, we can say that the above mentioned named formula is valid for all the chart types. If it makes sense, you can create other chart types using named formulas. The data will always appear as it was placed in one row. This way you have only one data series on the chart, so only one color. The array technique could be for simple cases too: you can easily and clearly visualize trend of data from 2D range. For example you have this data: If you create a column chart on the usual way, using the rows as data series, and set the column gap, you can have this chart: And this is how it looks like using only one data series as named formula: data=Sheet1!$C$3:$F$7*1 and also a named formula for axis labels: axislabels=Sheet1!$B$3:$B$7&" "&Sheet1!$C$2:$F$2 Simply the row and column headers are concatenated. The difference is more visible using a line chart. Using this chart you can not put the quarters side by side, only above each other. Using the same named formulas as above, it is possible to create a trend line: We received the below explanation from Jon Peltier of http://peltiertech.com  many thanks for the supplement! When you select the block of data in the Select Data dialog, select
the first row of the block, then hold Ctrl while selecting the rest of
the blocks.
The Series X and Y Values in the dialog will appear as: =(Sheet1!$B$3:$D$3, Sheet1!$B$4:$D$4, Sheet1!$B$5:$D$5, Sheet1!$B$6:$D$6, Sheet1!$B$7:$D$7) and =(Sheet1!$B$11:$D$11, Sheet1!$B$12:$D$12, Sheet1!$B$13:$D$13, Sheet1!$B$14:$D$14, Sheet1!$B$15:$D$15) and the series formula will look like: =SERIES("testing",(Sheet1!$B$ 3:$D$3,Sheet1!$B$4:$D$4, Sheet1!$B$5:$D$5,Sheet1!$B$6:$ D$6,Sheet1!$B$7:$D$7),(Sheet1! $B$11:$D$11,Sheet1!$B$12:$D$ 12,Sheet1!$B$13:$D$13,Sheet1!$ B$14:$D$14,Sheet1!$B$15:$D$15) ,1) 
Igor: What is this? Freddy: Schwarzwalder Kirschtorte Monster (OS): Mmm Freddy: Oh, do you like it? I'm not partial to desserts myself, but this is excellent Igor: Who are you talking to? Freddy: To you. You just made a yummy sound, so I thought you liked the dessert Igor: I didn't make a yummy sound. I just asked you what it was Freddy: But you did, I just heard it Igor: It wasn't me Inga: It wasn't me Freddy: Well now look here, if it wasn't you, and it wasn't you Monster (OS): Mmm. by The FrankensTeam Initially I was inspired by Peoplemovin Carlo Zapponi's chart. He describe his work with these words: "peoplemovin shows the flows of migrants as of 2010 through the use of open data (see Data Sources). The data are presented as a slopegraph that shows the connections between countries. The chart is split in two columns, the emigration countries on the left and the destination countries on the right. The thickness of the lines connecting the countries represents the amount of immigrated people." It is a very nice job that I wanted to reproduce in Excel ...and I created a file but did not get a satisfactory result (if you want I can send the file) ... sometimes we have to know how to lose ... or maybe we just need to change strategy. It is what i did :) i thought that I had to group the data. If I could not use the data so in detail perhaps I could show flows from one region to another. I downloaded Bilateral migration matrix 2010 file and worked a bit on it. I used the groupings regions as defined by The World Bank and in the end I used a flow chart inspired by Cosmograph Willard Cope Brinton's chart (published in the Chapter 8  Flow Chart 1939 in his Graphic presentation). Now the result is much more satisfying :) I used the mouse rollover technique (a small function VBA to be activated when opening the file) to highlight the flows, the chart is bilateral, you can select the region right or left and see migration flow to/from that the region selected. I had to fix the formulas in values (where it was possible) to lighten the file ... and probably could do much lighter and without all of that data ... but it came out so and maybe will improve it later :) Click here to download the Excel file. See also here: Anatomy of an Excel chart  Cosmograph dynamic version Talent traffic chart with chord diagram in Excel Life Expectancy by Nathan Yau's chart with Excel Edward Tufte's Slopegraphs in Excel 
Freddy: So then, what we're aiming for is a being approximately seven feet in height, with all features aither congenitally or artificially proportionate in size. Igor: Something like this. Freddy: Hello. You've caught something there. Crude yes, primitive yes, perhaps even grotesque. Yet something tells me that this might be our man. by The FrankensTeam There are two reasons to write this post. First of all, to share a good news we are really proud of: our dashboard won the Chandoo data visualization contest! We would like to say thank you to all those who voted  either to our or to the other 11 brilliant works! Take some time to analyze them  there is a lot to learn from the masterpieces. Second… The secret  Teamwork!
The beginning of the story was that IngaKris had seen a post somewhere suggesting to create a bubble+line chart combo starting from a normal line chart and adjusting the marker size manually. Because this way the result is totally static, she started to think if it was possible in a dynamic way. Bubbles could be done dynamically only on a bubble chart, so this must be the base chart. But bubble chart could not be combined with any other chart types, so we could only use the elements of the bubble chart itself. Scrolling over the “Add chart element” option, the trendline seemed to be suitable. 
Fraü Blucher: And this is your room. It was your grandfather Victor's room. by The FrankensTeam In the last post we shared a vintage chart reproduction found in a great, old book about charts. Today's vintage chart has a totally different story. The below picture is from a (Hungarian) wall calendar illustrated with photos of legendary actors and actresses. The chart is a stage prop... well, it is not suitable for practical use... it's for theater use only! Characters: Tivadar Bilicsi (Uncle Bezzeg), Kálmán Latabár (Dániel) and Kamill Feleki (Glauziusz) Here is the stage prop enlarged: The title says "Table of competition". The segments in the circle represents store departments, the lines are the performance of those departments in the socialist competition. The best department ("Férfi ruha" = Male clothing) is marked with roman number I. NOTE: In Excel 2013 it is possible to change the size of the data
labels individually, so the final result with the text labels looks much more better comparing to 2010 version. It was a funny exercise to create this chart though as you see, it is definitely not for practice, but some tricks could be handy in other projects:

Dr. Frederick Frankenstein: There's always a device. If I could just
spot the triggering mechanism. Hello. It seems louder over here. Hand
me that candle, will you? Put the candle back. Alright, I think i have
it figured out now. [...] by The FrankensTeam When a new chart is created, the order of the data is the same as on the worksheet. In many cases it would be helpful to see the data in ascending or descending order. In this article we would like to show you how you could avoid sorting the original data, and create a dynamic, sortable chart using named formulas. The most convenient way to choose the sort direction is a validation list. You can choose no sort, ascending sort or descending sort and the chart will change immediately. Something like this: You can download the sample file used in the above explanation. Let’s start with this data table: First of all, we will use dynamic named ranges to make possible adding new data to the data table. Define the name "values": =data!$B$2:INDEX(data!$B$2:$B$100,COUNTA(data!$B$2:$B$100)) and "labels": =OFFSET(values,,1) Formulas to sort the data Sorting values with formula is not difficult with SMALL and LARGE functions. These formulas return the kth smallest or largest data in a data set. We will use these as array formula, with a vector of integer numbers like: {1,2,3,4} as second (k) parameter, so the result is the sorted array of our values. For the vector of integer numbers ROW(values)MIN(ROW(values))+1 could be used  it will result as many numbers as many data we have.Define two names: Sort_values_asc: =SMALL(values,ROW(values)MIN(ROW(values))+1) Sort_values_desc: =LARGE(values,ROW(values)MIN(ROW(values))+1) Formula to choose from the sort directions First, create the validation list in cell E2 with this list: "nosort","ascending","descending". Here you will set what kind of sorting you want to see. For the chart, we need one data serie, so we need a formula wihch will provide the data set according to our choice. Excel’s CHOOSE formula will help doing it: based on an index value this formula can choose a name from the list of names. In other words, you choose ascending and CHOOSE formula will give the name Sort_values_asc back. Because CHOOSE formula needs an index number, we have to use MATCH to determine which element of the list was chosen: MATCH(data!$E$2,{"nosort","ascending","descending"},) will result 1, 2 or 3 according to our choice. Now we can build up CHOOSE using this matchresult as first parameter: =CHOOSE(MATCH(data!$E$2,{"nosort",
So if your choice in E2 is AZ, MATCH will result 2, and CHOOSE will give back Sort_values_AZ named formula. We can go one step further and use always “nosort” if nothing was chosen. We defined the name for_bar_serie: =CHOOSE(IF(ISBLANK(data!$E$2),1,MATCH(data!$E$2,{"nosort",
Formula to sort the labels We have to use a support column on the sheet to put the labels to the same order as the data on the chart. Because there could be duplicated values, the formula must be prepared to bridge this problem. Here it is: =INDEX(labels,MATCH(1,(INDEX(for_bar_serie,ROWS(S$2:S2))=values)*(COUNTIF(S$1:S1,labels)=0),0)) Confirm with Control+Chift+Enter and copy down. You should copy it more cells then you have in the data table, so you will not need to change it when you have more data. The most important part is this: (COUNTIF(S$1:S1,labels)=0) where we check if the label is alrady used in the list above the cell. In case of duplicated values this part makes possible not to choose the same label twice. Create the chart Now you can create the chart. Simply highlight the values on the sheet, and insert the chart you would like to use. Then go to the Select data… (after rightclick on the chart) and overwrite the range with the named formula for_bar_serie created above: Then edit the Horizontal (category) axis labels by selecting the range of the support column (instead of the original labels!). You should select a larger range than the original data table. You are done! 