Today, let’s travel in time. Pack your photon ray guns, extra underwear, buckle your seat belts and open Excel!
Of course, we are not going to travel in time. (Come to think of it, we are going to travel in time. By the time you finish reading this, you would have traveled a few minutes) We are going to learn how to travel in time when using Excel. In simple terms, you are going to learn how to move forward or backward in time using Excel formulas. So are you ready to hit the warp speed? Let’s beam up our Excel time machine. Tip 0 – Date & Time are an illusionMost important tip for Excel time travelers is to understand that Excel dates & times are just numbers. So when you see a date like 17-October-2013 in a cell, you can safely assume that it is a number disguised to look like 17th of October, 2013. To see the number behind this, just select the cell and format it as number (from Home ribbon).Now that you understood this concept, let’s jump in to the 42 tips. All these tips assume a date or time value is in the cell A1. Staying at present:
Going back in time
Going forward in timeWe, time travelers are smart people. Once you know that turning the knob backwards takes you to past, you know how to go to future. So I am giving very few examples for going forward in time.
Finding the amount of time traveled
Fixes for common time travel hiccups
Quiz time for time travelersI see that you safely made it here. I hope you had a good journey. Let me see how good your time traveling is. Answer these questions:
Building your own time machine? Check out these tips tooIf you work with date & time values often, then learning about them certainly pays off. Read below articles to one up your time travel awesomeness.Good luck time traveling. I will see you again in future |
Thursday, October 17, 2013
42 tips for Excel time travelers
Saturday, October 5, 2013
Formula Forensics No. 035 Average the last 3 values greater than 0
A couple of weeks ago Amanda asked a question at Chandoo.org
“I need to calculate a moving average of the last 3 months.
However, if one of the months contains 0%, I want it to ignore that and take the last month that didn’t have a zero.
For instance, my data is this: April = 100%, May=0%, June=97%,July=98%, August=0%.
My formula is only looking at June July and August, but since August has a 0 I would like it to look at May June and July.
And since May has a 0, I would like it to look at April, June and July.”
I offered a solution which is an array formula.=AVERAGE(AVERAGEIFS($B$3:F3,$ Today I am going to try and explain what it is doing and how it works. As always at Formula Forensics you can follow along with a sample file: Download Here The ProblemHow do we write a formula to extract the last 3 non zero values?What if there is more than 1 zero value? What if the zero values are non-contiguous? This can all be shown by: Average of the last 3 records – No Zeroes Average of the last 3 records – One Zero in Current Month Average of the last 3 records – One Zero in a Previous Month Average of the last 3 records – Multiple Zeroes A Solution=AVERAGE(AVERAGEIFS($B$3:F3,$Normally at Formula Forensics we start in the inside of a formula and work out, but today we are going to start at the outside and work our way in. The solution: =AVERAGE(AVERAGEIFS($B$3:F3,$ This function is the Average() of an Averageifs() function. What is going on here you ask ? We know that the Average() function will average the constituent numbers. eg: =Average(6, 8, 10) = 8 But doesn’t Averageifs return a single number? The average of its components! We’ll mostly, but not always. Lets have a look: If we remove the outside average and just look at the inner Averageifs() function In the sample file, Cell F15 you will see: =AVERAGEIFS($B$3:F3,$B$2:F2, Excel evaluates this as: ={0.5,0.9,0.7} So the Averageifs() function is returning 3 values being 0.5, 0.9 & 0.7 These are the last 3 values greater than 0 ranked from latest to earliest by date Which is exactly what Amanda asked us to average So we will need to look inside the Averageifs() function to see what is going on. The Syntax for the Averageifs() function is: AVERAGEIFS(average_range, criteria_range1, criteria1, [criteria_range2, criteria2], …) in our example Average_range : $B$3:F3 This is the Values we want to average from teh start of the data up to the current cell Criteria_range1 : $B$2:F2 This is the Date Range from the start of teh data up to the current cell Criteria1 : LARGE(IF($B$3:F3>0,$B$2:F2),{ So in in English, the Averageifs function is being asked to return a single value from the Value row where the Date row matches the largest, second largest and third largest criteria. In each case Averageifs will average the number but as it is a single number it returns the value by itself. Lets now step into the Large() function and see what is going on. LARGE(IF($B$3:F3>0,$B$2:F2),{ The Large() function has the syntax =Large(Array, k) In our example: Array: IF($B$3:F3>0,$B$2:F2) k: {1,2,3} This is an array and hence asks for the Largest (1), Second largest (2) and Third Largest (3) values So we are getting the 3 largest values from the formula: IF($B$3:F3>0,$B$2:F2) What is the IF($B$3:F3>0,$B$2:F2) formula doing? If you put the formula IF($B$3:F3>0,$B$2:F2) into a blank cell say F31 and press F9 Excel returns: {41275,41306,41334,41365, These are the date values of the date row up to the Column we are working in You can see these in Row 1 of the sample file You will notice that there is a False value in the position of the Column which has the Value of 0% This is derived by the If() Function. The If() function is saying: IF($B$3:F3>0,$B$2:F2) If the Value Row >0 ($B$3:F3>0), return the Date Row ($B$2:F2) else return False The False isn’t shown in the formula, it is returned by Default when a False argument doesn’t exist The standard Syntax for If is =If(Criteria, Value when true, Value when false) In our case we don’t have a Value when false component and so Excel simply places a false in as the answer. Applying the formula using Ctrl+Shift+Enter forces Excel to Evaluate the formula as an Array Formula What this means in practice is that the Formulas are evaluated 3 times as per the array {1,2,3} effectively extracting the last 3 values that match the last 3 dates where the Value is >0 What if I want to Average a Different Number of Days?You have two choices1. Change the Array ManuallyIf say you want to average the previous 5 valuesYou can modify the array manually =AVERAGE(AVERAGEIFS($B$3:F3,$ This is ok if it is not done regularly or is only slightly different to the existing array. But if you want to setup the top twenty you need to type {1,2,3,4,5,6,7,8,9,10,11,12, Which is quite clumsy! Bad Hint: You can copy the array from above. Or you can automate it using the technique below. 2. Insert a Formula to Setup the ArrayYou can add a small function that will automatically setup the array like:=AVERAGE(AVERAGEIFS($B$3:F3,$ This assume that cell F39 has a number which is the number of periods you want to average You can read more about how the ROW(OFFSET($A$1,,,F39,1)) part works in previous Formula Forensics posts DownloadYou can download a copy of the above file and follow along, Download Sample File.Other Posts in this SeriesThe Formula Forensics Series contains a wealth of useful solutions and information.You can learn more about how to pull Excel Formulas apart in the following posts:http://chandoo.org/wp/formula- Two Challenges1. Can you solve this another wayJust after I posted my solution, Chandoo posted an alternative solution which you can read at:http://chandoo.org/wp/2009/04/ Can you solve this problem another way ? Let us know in the comments below: 2. Your Challenge?If you have a clever formula and would like to become an author here at Chandoo.org please consider writing it up as I have done above. Alternatively you can send the formula to either Chandoo or Hui. |
Subscribe to:
Posts (Atom)