I just finished a project where I was using IPython and Pandas a lot more than spreadsheets. I found that after using Pandas, I started to think about data a little differently. When I returned to Google Spreadsheets, I realized it wasn’t capable of displaying the data in the way that I wanted to. So I took another look at the Google Spreadsheet SPARKLINE function and found that, in many cases, they provide a more informative and much quicker visualization that charts. Once I got started, I figured I would exhaust the possibilities of the feature and share them here so that you can get started with them quickly and easily when you need to.
There are four different types of sparkline charts. I have a section for each:
Horizontal Bar Charts
Single Series Data
Best for: Comparing every cell in a column
Basic Bar Chart
=sparkline(A36,{"charttype","bar";"max",100})
It’s pretty straight forward. Make sure to set the max option so that the bar does not take up the full column.
Reverse Direction with Max Argument and Color
=sparkline(E42,{"charttype","bar";"color1","teal";"rtl",true; "max",max(C$36:C$44)})
rtl (right to left) changes the direction of the bars. Use the max function for the value of max with the data column as the range to set the width limit to the highest value in the data column.
Stacked Series Data
Best for: Comparing multiple columns with each row as parts of whole
Stacked “Versus” Chart
=sparkline(E49:F49,{"charttype","bar";"color1","EB4967";"color2","73A4D3"})
Set color to hex codes for to make visually appealing sparklines. No value for max will set the width of the bar to the width of the cell.
Stacked Bar Chart
=sparkline(E58:H58,{"charttype","bar";"color1","A7DBD8";"color2","69D2E7";"max",max(I$58:I$61);"nan","ignore"})
The Q1 and Q2 values for the 2014 row do not have numerical values. Setting nan to “ignore” will exclude these values from the sparkline bar chart. The max option is set to the highest value of the Total column.
Line Charts
Single Series (Y) Data
Best for: Displaying time series data where value change over consistent time intervals
Standard Line Chart and the Effects of ymin
=SPARKLINE(B3:G3,{"charttype","line"; "color","indigo"; "linewidth",2})
=SPARKLINE(B3:G3,{"charttype","line";"color","indigo"; "ymin",0})
This chart uses a custom color and linewidth. It also demonstrates how to change the chart’s background color by changing the cell’s background color. Both charts display the same data, but because it sets the ymin option to 0, the chart looks much different. Be careful when setting limits and cell dimensions. These adjustments can make the data appear to say different things.
Line Chart with Conditional Coloring
=SPARKLINE(C7:H7,{"color",if(H7>C7,"green","red");"ymax",100; "linewidth",2})
The color is set to an if formula. This formula compares the first and last value of the series, and if the last value in the series is greater than the first value, the line color option is set to green; and if not, the color is set to red. The slope formula is also good for this, but it is better with X,Y series data.
Dual Series (X and Y) Data
Best for: Displaying time series data where value change over inconsistent time intervals
Line Chart with Conditional Coloring
=SPARKLINE(C12:H13,{"color",if(slope(C13:H13,C12:H12)>0,"green","red"); "linewidth",2; "ymax",25; "xmax",15})
There is a lot going on here, but it builds on the previous example. The slope function is used to set the color for the line. If the slope is greater than 0 the line is green; else it is green. Additionally, ymax and xmax are set to values greater than any data value. This creates a margin around the line chart.
Line Drawing… Because you can
=SPARKLINE(C26:O27,{"color","red"; "linewidth",8})
The lines will follow the X and Y coordinates even if they are not consecutive. Go nuts!
Win/Loss Charts
Best for: Displaying binary data like wins and losses
Custom Colors and Colored Axis
=SPARKLINE(B$2:G$2,{"charttype","winloss";"color","teal";"negcolor","grey";"axis",true;"axiscolor","black"})
Colors help draw attention to the data more quickly than just shapes.
Custom Colors and Colored Axis
=SPARKLINE(B$2:G$2,{"charttype","winloss";"color","grey";"axis",true;"axiscolor","black";"firstcolor","teal";"lastcolor","yellow"})
The firstcolor and lastcolor options are a good chance to use custom coloring. These can be helpful for summaries of a series of data like whether a team was over or under .500.
Custom Colors and Colored Axis
=SPARKLINE(B$6:G$6,{"charttype","winloss";"color","grey";"axis",true;"axiscolor","black";"nan","ignore";"empty","ignore"})
The nan and empty options determine how non-numeric or empty data is treated. In this case, the first cell is empty, and the fifth cell is text. Both cells are ignored.
Column Charts
Best for: Comparing a series of values against each other
Dealing with Negative Values
=SPARKLINE(B$2:G$2,{"charttype","column";"color","teal";"negcolor","purple"})
=SPARKLINE(B$2:G$2,{"charttype","column";"color","teal";"ymin",0})
Column chart sparklines offer two ways to call out negative values. The first is to set the negcolor value and the second is not to display them altogether by setting the ylim value to 0.
High and Low Value Accent Colors
=SPARKLINE(B$2:G$2,{"charttype","column";"color","teal";"negcolor","purple";"lowcolor","fuchsia";"highcolor","aqua"})
The highcolor and low color options make is easy to identify the high and low values in the data series.
Bonus: Hex Color Options
Google Spreadsheets is great in that it allows you to customize alot of the visual display. Here are a few nice color palettes to make you data a bit more visually pleasing. For more, checkout: http://www.colourlovers.com/palettes
#69D2E7 #A7DBD8 #E0E4CC #F38630 #FA6900
#41F4E3 #73A4D3 #F0D9A8 #70109 #EB4967
#00A0B0 #6A4A3C #CC333F #EB6841 #EDC951
#ECD078 #D95B43 #C02942 #542437 #53777A
I hope that covers any visualization that you might need to use. Let me know in the comments if you are having any trouble. Tomorrow I will publish a post on how to make funnel diagrams using the Sparkline Function.
Hi, thanks for posting all that great information!
I have just seen a good extra tip:
“When you address the option like an object, then you should use a backslash: =SPARKLINE(C11:N11;{“charttype”\”bar”}”
http://stackoverflow.com/questions/28654796/formula-parse-error-in-google-spreadsheet-sparkline
Best!
Hi Everton, thanks for adding to the post. I have not had this type of formula parse error but I have a feeling your comment will help people in the future. Thanks!
hi thank you for the post. Is there any possibility that we can draw a candlestick chart in a single cell like you explained for line chart. Ofcourse we need OHLC price. we can get it in google spreadsheet. so is there any possibility? Thanks
Hi Jo, Unfortunately, there is no candlestick chart for Google Spreadsheet Sparklines per se. You might have to get created with the options above. Sorry…
Hi, Thanks for this post!
I still need to add a tooltip over the sparkline, so do you have a way to achieve that?
Thanks
Unfortunately, to my knowledge, there is no tooltip feature for Google Spreadsheet cells.
Thank you! Quick question – what if you use a sparkline to show progress in times taken to swim a certain distance. As the times get faster (less) the sparkline will appear to fall. This can be confusing for the viewer – is there an easy way to switch the direction of a sparkline so that it will appear to climb as the times get faster?
Here’s the spreadsheet where I need to configure it.
https://docs.google.com/spreadsheets/d/1COlZwXumIPFSgBvX3zWDEn0e7SxUB1IDhursTL4n52Y/edit?usp=sharing
I’m sure it’s a really easy formula tweak – thanks if you can help!
It seems that all you should have to do is pick an arbitrary lap time limit (a really slow time) and subtract the swimmer’s time from that limit. As the swimmer’s time improves, the number subtracted from the arbitrary limit will decrease, leaving the remaining number to increase- creating the effect of a line that goes up instead of down.
I wonder if, when it comes to sparklines, we can go actually about referring to cells by alternate columns (say, B1, D1, F1, for instance) instead of the usual “B1:F1”.
Ops! Found it: by using an array {B1, D1, F1] everything worked like a charm!
Nice find! Was just about to say…