Spreadsheet magic

I once stumbled upon a quote “if people knew how to use grep, awk, sed, and xargs, half of the applications would never need to be created.” I think roughly the same can be said about spreadsheets - knowing how to use just a few features and functions allows handling quite complex use cases within just a few hours. In this post, I’ll cover a few of those most-useful-but-a-bit-complex spreadsheet’s functions and features and show how they combine to build a “habit tracker” spreadsheet.

What we are building

First, let’s cover the use cases for the habit tracker. Long story short, a habit tracker is an app (or tool) that helps develop (or quit) a habit: “read at least 30 min a day,” or “quit smoking.” Habit trackers support inner motivation to do so with “external” reinforcement:

  • rewarding good behaviors via “green tick “effect” - ticking a checkbox or crossing an item grants some sense of accomplishment
  • reducing perceived “toughness” of the task - “I will do this for two months” is a lot harder than “I will do this today” for the same two months.

Here are the use cases our habit tracker sheet would need to fulfill:

  • Support habits that need to happen every day, or only on some days of week/month.
  • Support qualitative and quantitative habits.
  • Allow setting goals for habits - for example, “run 20 km a month” or “read for 30 min every day”.
  • Clearly show current progress towards the goal and remaining “distance.”
  • Enable “green tick” and “do today” effects.

Spreadsheet

Here’s the result - Habit Tracker spreadsheet. I dare to assume using it is quite self-explanatory - add/remove the goals by inserting/removing columns on the Template sheet, putting the weekly schedule onto the Technical sheet, adjust conditional formatting if something becomes off1, copy the Template sheet, set the day you want to start, and off you go.

 Screenshot showing a partially filled habit tracker spreadsheet for January 2022. The date on the screenshot is January 8th, 2022. The first row contains headers: Date, DOW, Day of Week, and finally Habit 1 to Habit 7. Date column contains dates starting with January 1st, 2022. About 20 dates are visible. The following two columns contain days of week numbers (DOW column) and names (Day Of Week column) Habits 1, 2, and 3 have a weekly schedule. 1st and 3rd habit have "days off" on weekends - the corresponding cells have blue fill and letter N. Habit 2 also has Tuesday and Thursday off, with similar cell contents and formatting. Cells with any value except N have a green color fill. Empty cells in rows 2022-01-07 and earlier have a red color fill; empty cells in rows starting with 2022-01-08 have no color fill. Habits 4 to 7 are "everyday" habits and have checkboxes in the cells associated with them. If the checkbox is marked, the cell has a green fill. Otherwise, cells that are "in the past" have red fill.

Now that we have a result in view, let’s finally talk about technical stuff - features and formulas used to build it. I’ll omit the very basics (like cross-tab references, references with fixed row/column, well-known functions, etc.) and focus on the most powerful and complex to use. All cell references are on the Template tab, unless the sheet is explicitly specified.

Functions

Sequence

SEQUENCE function - fills the specified number of rows and columns with sequential values with a specified step.

Example: SEQUENCE(1, 5, 0, 2) - creates a (1-column x 5-row) block with [0, 2, 4, 6, 8] values

Used at cell A7 to automatically create a column of dates in the selected month.

HLOOKUP & VLOOKUP

VLOOKUP(search_key, range, index, is_sorted) is a powerhouse behind many tricks that look
magical. The simplest way to explain it is to imagine that the range becomes an isolated sheet. First, the index argument selects a column from that “virtual sheet.” Then, search_query is looked up in the first column of the “virtual sheet.” If VLOOKUP finds the match, it returns the value of a cell that resides at an intersection of this row and the selected column. is_sorted makes this lookup “inexact” and select the row with the nearest match - the largest value that is less or equal to the search key.

HLOOKUP does the same, except it fixes the row and searches in the first row to select a column.

Example:

  • VLOOKUP(A10, Technical!B7:D25, 3, FALSE) - looks up exact value of A10 in Technical!B7:B25 and if found returns the value of a cell in a D column

Used at:

  • Cells C7:C37 looks up weekday name by weekday number2.
  • Cells D7:F37 use VLOOKUP in conjunction with IF to apply the weekly schedule to the goals.

INDIRECT, ADDRESS, ROW & COLUMN

INDIRECT returns the cell’s value by the address. Simple put INDIRECT("B2") becomes the value of B2. It might look redundant at a glance (=B2 does the same, and is much simpler), but INDIRECT starts to shine when used with an alternative reference mode INDIRECT("R1C1"; FALSE) or in conjunction with ADDRESS.

ADDRESS takes row and column umber numbers and returns the cell address.

ROW and COLUMN simply returns the row and column of a cell, both 1-based3. Calling them without arguments return the values of a current cell.

Examples:

  • INDIRECT("RC[-1]"; FALSE) - returns the value of a cell to the left.
  • INDIRECT(ADDRESS(ROW()+2; COLUMN();;;"Technical")) - fetch value from a Technical tab and 2 rows below. E.g. C2 becomes Technical!C44.
  • INDIRECT(ADDRESS(COLUMN(); ROW();;;"Technical")) - fetch value from a Technical tab and “transpose” the address
    E.g. C2 => Technical!B3, D7 => Technical!G4. Quite handy if “values” on one tab needs to become column headers on the other (yes, similar to pivot table).

Used at:

  • Limited use of COLUMN in cells D7:F37 - practically could’ve done without it.
  • A lot of uses for conditional formatting, which we’ll cover in the corresponding section.

“Open” range references

Range references are widely known and don’t need explanation. However, there are less-known reference styles that can come in quite handy.

Examples:

  • A7:A - A7 and rest of the A column. Similarly, C7:7 - entire 7th row, starting with column C
  • B12:G - columns from b to G, starting with 12th row.

Used at:

  • Cells D3:D7 use D7:D to sum all the records in the column.

Conditional formatting

Conditional formatting makes the dry and terse spreadsheet(s) speak to the emotional structures in the brain. With conditional formatting (and a little we can finally make the “green tick” effect work. Two sets of conditional formatting rules exist in the tracker sheet: one for “habits with a weekly schedule” and the other for “everyday” habits.

Note: conditional formatting rules are evaluated sequentially and terminate after the first matching condition.

Weekly schedule ruleset:

  • If cell value equals N - blue fill
  • If the cell contains data - green field
  • Custom formula “if a date is in the past” - red fill.

Every day ruleset:

  • If equal to True - green fill
  • Custom formula “if a date is in the past” - red fill.

The “if a date is in the past” formula is probably the most complex thing in this entire sheet. Here goes:

=AND(
  NOT(ISBLANK(
    INDIRECT(ADDRESS(ROW(); 1))
  ));
  TODAY()>INDIRECT(ADDRESS(ROW(); 1))
)

In plain English: “if the first cell in the row is not empty and its value is smaller than today’s date.”

Recap

Here’s how the various features listed above contribute to achieving the use cases we listed.

Use caseFeatute
Weekly and every day habitsVLOOKUP + WEEKDAY
Qualitative habitsCheckboxes
Quantitative habitsDon’t need special support
Setting goals, tracking progressRow with goals + SUM with “open” range
“Green tick” effectConditional formatting + INDIRECT + ROW
“Do today” effectSEQUENCE + Conditional formatting + INDIRECT + ROW

Extras

Other spreadsheet features you might want to have in your toolbelt:

Pivot table - a very flexible and powerful instrument to aggregate data, and even more so if used in conjunction with VLOOKUP. Record what you eat and how much, VLOOKUP nutrition data, aggregate per day with pivot table, pull onto a separate tab, add target values via VLOOKUP again, compare actual with targets - here’s your diet monitor5.

Relative references - also know as RC-format. Only work with INDIRECT, but can be extremely handy for sophisticated conditional formatting scenarios. For example:

  • fill red if cell to the right is larger than B2 - =INDIRECT("RC1"; False) < B2
  • fill green if current cell is larger than 0,5 * B2 - =INDIRECT("RC"; False) > 0,5 * B2

Conclusion

This spreadsheet has almost complete feature parity with one of the paid habit tracker apps - and only took 2.5-3 hours to create. Mastering just a few spreadsheet features (and googling the others) allows supporting many use cases that otherwise require a dedicated app. Moreover, you get to customize your solution as you please, export in virtually any format you like, share, comment, collaborate, etc., etc., etc., out of the box.

Practically, even moderately complex spreadsheets can cover more complex scenarios from a variety of areas - financial modeling, project prioritization, evaluating job offers, stock picking, and many more.

  1. Common source of issues is moving the content - spreadsheets are not always intelligent enough to adjust formulas and conditional formatting when source data moves. However, adding/removing columns is more robust, so try to use it as much as possible. ↩︎

  2. This could be achieved without VLOOKUP by text formatting TEXT(value, "dddd")… Unless you need names from a different locale/language ↩︎

  3. i.e. for cell A1 both ROW and COLUMN return 1, not 0. ↩︎

  4. This can be easier achieved by using a reference + “pulling” the cells ↩︎

  5. the challenge here is to obtain nutrition information and carefully record what you eat. The rest is pretty mechanical. ↩︎


© 2021 Eugeny Kolpakov. All rights reserved.

Powered by Hydejack v9.1.5