Scatter chart, with one text (non-numerical) axis

Question

Lets imagine I manage 100 pizza stores. I want to plot a chart each day, of how long deliveries are taking.

My data might look like this:

London      23
London      22
London      44
London      18
Paris       34
Paris       14
New York    45
etc

I would like a scatter plot; numbers (minutes) on one axis, and cities on the other axis. I would also like each city to appear in the order I choose.

I can only get this to work if I substitute each city name for a number (city number), but that has the effect of the axis labels being the city numbers (rather than city name). How can I substitute the number for the real city name? I do not want to label each data point, only the city axis! I've tried a secondary axis, but again am unable to get the city name as the label.

I'd like to use 'simple' Excel to solve this, but if there's another solution (e.g. VBA, Python), that'd interest me too...

Answer 1

Excel scatter plots cannot take names instead of values on their x-axis. They assume a number series for the x-axis. If you want to replicate the effect of a scatter plot but use named x-axis values, you can use a line chart instead.

First, though, you are going to need to re-order your data. Each data point for a city must be in a separate column on the same row. For your example data set:

   A         B   C   D   E    
1  London    23  22  44  18
2  Paris     34  14
3  New York  45

Select the data and choose a line chart style with data points marked. Double-click a line to go into the Format Data Series dialog. Under Line Color choose No Line, and click OK. You can repeat the process quickly with the rest of the data set by clicking on a line and pressing CTRL+Y.

The end result will look something like this:

Answer 2

I ended up using PyPlot, with the excellent guidance of http://matplotlib.org/gallery.html, to see the features that I was interested in using. Not for the non-programmer, but I think PyPlot can do absolutely anything.

Answer 3

It does, you just need to choose a two-line plot instead of a one-line plot.

Answer 4

For people with the same problems, there is a usefull tool named FunFun, which allows people to use web languages like Html, Css and JavaScript in Excel.

Javascript has a lot of powerful libraries for data visualization, like Charts.js and D3, that you can use with this Excel add-in to create any chart you want.

I made this chart with Chart.js on the FunFun website that I directly loaded in Excel by pasting the link of the funfun online editor below:

https://www.funfun.io/1/edit/5a3be7c2b848f771fbcdebbc

As you can see, you can put the city name correctly in the order you want just by changing the code on the Funfun online editor witch has an embedded spreadsheet, where you can see the output of your code instantly.

Here are some screenshots(Website to Excel Add-in):

Hope this helps !

Answer 5

You can assign a number to each of your x-axis text categories. For example: North = 1 East = 2 South = 3 West = 4

The data points will appear in a row above each number on the x axis.

Answer 6

This should be a High-Low-Close stock chart. If you're measuring averages based on multiple delivery times you can show the average time, the lowest time, and the highest time. With a little bit of cell movement and averaging, you can average the highest third times, the lowest third times, and the middle third times and show the averages of each in the HLC chart. With a large enough number of measurements and a little analysis, this would net a pretty good result with regards to managing driver/store employment, delivery grouping, and delivery area management.

Add in an averagetimeofdelivery/priceoforder measurement over several months and you could then project future earnings based on delivery area. For instance, if your high profit deliveries are less than 10 minutes away, and you get a lot of deliveries that are 15+ minutes away but net small profits, then based on $ saved/profitloss you could possibly make some changes to your delivery area and then driver count. Paying fewer people is a great recipe for increasing profits. I know that's much farther in the weeds than the post is going, and it's a super old post but I used to deliver pizzas and had a similar excel problem (unrelated to pizza delivery), and I felt like ranting.

Answer 7

I have a kind of workaround. Add the labels you need in column 2 (second column of the table). Create the smooth scatter chart and right click the curve and click "Add data labels". Right-click one of the data labels next to the curve and click "Format data labels". Untick x-value, untick y-value and tick "Value from cells" and select the relevant range (Second column in my case, but can be anywhere). It is not a perfect solution as you still will have numbers showing on the x-axis (or show labels set to "none").

Answer 8

This is an old thread, but I was looking for a way to use a Scatter Plot. In Excel you can just use a line chart, as indicated previously (thanks!), set your Markers as desired and no line. I added a couple of data point labels. I pivoted hour of day as text, using: HOD =TEXT(HOUR([@[Call Time]]),"00") with HOD on rows, DOW on Columns. In the "Select Data" for the chart, put in each day with its hourly volumes. Note, to ensure my data customer understood, I added a heat-map of the pivot with the volumes beside the chart, thus providing two means of viewing the high-volume days/times. Hope this is useful to someone!

Volume of Calls by Hour of Day and Day of Week

The Accompanying heat map

Answer 9

There is a way to build a scatter chart even with both axes formed by categories but it is slightly tricky and requires a hybrid solution. However, the effect is good and you can then take a screenshot of the graph to reproduce it, for example, in a PowerPoint presentation.

Let us suppose to have a table that contains the number of stores in different European cities. We provide two pieces of information for each store: location (city) and type of store. For example:

The first step is to duplicate the table and use integer numbers instead of categories, as follows:

Second step, insert a bubble chart with the numeric categories on X and Y axis, and number of stores as bubble size, as follows:

Third step:

1. remove title, axis, legend, and any other chart element but the chart area,
2. enlarge the chart area to full size,
3. remove chart area and chart border,
4. resize chart area to align to sheet cells,
5. use cells on top, bottom, left and/or right of chart to add categories

The result is:

Enjoy! :-D