Skip to content

alejandrohagan/learningR

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

 ---
title: "Readme"
format: gfm
---

This is a draft of my book "Bridging the Gap: Learn R Alongside Excel - a Survival Guide for Corporate Environments" which documents my 10 year learning journey of business intelligence.

Why do we need another book about R or Excel?

While there are amazing books about R, specifically in the [www.bigbookofR.com](www.bigbookofR.com) I have found that they either

-   Tend to be overly focused on statistics or theory where the application to the Corporate environment may not always be immediately understood
-   Are too beginner focused that do not take into account the challenges of data reporting in a Corporate environment


# Mission and Vision:

After reading this book you will:

-   You will know how to apply tools to the real life situations that you have faced with an emphasis of effectiveness and not theory (as defined by your team & maintenance of the reports you build)

-   We will cover some introductory concepts however, the focus is mainly on application but we will direct you to resources for additional learning

-   There will be better ways to do things that we are teaching you however, they often involve wider resources or time commitments that you may not be ready for

# Introduction

I've scoured every tip, every yahoo message board, every stackoverflow/ mr. excel, . I've read all the XX for dummy books, blogs, secret forums. Youtube videos for days. Udemy? Done it. XXX I know it. I've reached all the dead ends. I've stared at the screen frustrated. I've been in your shoes as we data increases yet tools get more complicated. Promises made and promises broken.

But I've cried, I've sweat blood and cried tears so that you don't have to. After working years in financail plannning, budeging , analysis, manipulating, system data, external data, data meant to help that hurts, non standard data, excedl models that are so large you pray that they

But there is a solution. If you are dreading your monthly report, your weekly stewardship and KPIs, I am here to tell you that this book will get you answers. Answers that you can use today to fix your problems. No complex macros to learn, no advanced languages. This will solve 80% of your problems, and will point you towards resources to solve the remaining.

This is meant to quickly refrence guide. I will teach you patterns techniques so that you can quickly and efectively get to solutions to yoru prblems. This means building up your analysis frameowrks and tool kits.

If you’ve been coutning, you've noticed I've used "effective" almst 15 times. What do I mean by that, I'm balancing your time, With all the skills, you can (and some of you should) signficantly advanced your knowledge but what happens? For many of you, you'll build a tool only you can use, you will need to spend signficantly amoutn of your time learning these skills and testing these skills. Things will go wrong. You will be frustrated, eventually you'll get it right, you will be better for it. But then you will move, the person replacing you will have no clue what you are doing and will quickly undo it, and sure you've learnt an advance skill, but unless your organization is build around that level of skill standard - you may not intend it but you will do more harm than good.

It also means you will be able to do a typical problem in an defined amoutn of time. Trust me, there is no point reading this if you can't do a simple analysis in 15 minutes. Why? Because you spent time training on to have any benefit with your work.

By the time you are done, you will gladly look forward to data. It will not intimidate you. You will be excited to quickly semi automate many of your processes. You will be confident. You will be focus on value and process.

# My assumption about you

-   You regularly use excel for most of your reporting, visualizations and presentations
-   You regularly inherit other people's spreadsheets that are challenging to maintain, update or change
-   Requests for additional information are me with 

# Book structure

## Vocabulary that is helpful to learn

File Type 

(excel, datasource, txt, csv)


Tables vs. Not Tables

Basically, you need ito start thinking of your data in tables, this will seriously unlock your potentials and make your data much easier to work with

Data Structure:

Pivoted vs. unpivoted


Consistency of data?

Are you new columns added? Will new categories be added? Will categories be split open

Granularity of inputs/outputs

You are going to quickly learn how confusing data can be with granularity, even simple questions of

Column types: Text, numbers, dates, mixed, etc

Excel makes this mostly easily for you but the reality is ou need to be explict in a column type

Variability of inputs / outputs

Does your data have inputs or sensitivities? Do you need

Size

Data is big and can quicky overwhelm. This will become painful true if your orgnaization is using excel for everything. Get ready for megasize files, that link to megsize files that link further to megasize files. No joke. Excel files get currupted, links are broken. Your life will suffer. Sometimes thee isn't a way out of that but sometimes there is.

Convention: Naming, colors, fonts

Color convention

# High level framework to think about data

ETL+C+V

This part is pretty crucial for you to start to understand as you will think through your process better. It will help you to manage others that send you data or that you send data too. Many of your data will have cross functions inputs. You will need to send out inputs, receive inputs, plug those, document those, have a way to track changes, etc

## Extract:

This is often the most critical part. It’s the part that can make your life incredible easy and incredibly simple

Consistency of data

Structure of data

Variability of data

## Transform

The more you get E right the less you need to do with T, but that being said, there is some you may need to do, adding columns, unpivot / pivoting, changing data types, grouping,

## Load

In this context, it will be either excel, connection only or data model but it could mean loading it into a cloud system or other input files

## Calculate

Now you do your actual calculations on your data, this can be in a structured template, or custom formulas, but this is when you will use either basic or advanced formulas

## Visual

This is how you take complex, confusing, granular information and make it understandable and easy to understand. This is where your learnings are applied

Application

Data Structure: Pivot vs. Unpivot

Pros & Cons

Pros

Easily see information

Which one is easier to find and understand

Identify patterns

Cons

Terrible to use in analysis

Information may already be aggregated and you loose granularity in what happens

May identify patterns but not the ones that we need

If you connect it to, very difficult to automate

Toolkit

Advanced formulas, edit connections

Power query to transform

Understand why this is important, lets take a simple example side by side

Two datasets, side by side, exactly same information just structured differently

Try to find the answers to these questions, using your existing excel skills

Unpivoted Data

Tables vs. Tibbles

How to add data

Adding column

How is this different?

It applies against all rows

If you reference a column, it references all columns

If you want a subset, you will need interim columns to aggregate or dissect data in separate columns

Adding rows

Generally automatically puts into table, we will see how this is useful

Form

How to make data entry easier

How to interact

Filter

Basic to advanced options

Sort

Basic to advanced options

Add slicer

Caution - this can be useful addictive and then super frustrating

Reference tables

Basic:

Naming tables / headers

Try to keep with some consistnecy, you'll see why this is useful

Here's a useful sheet to help

Writing formulas against tables

Bucket your Formulas

AGGREGATION

LOGIC

REFERENCE

STATISTICAL

MATHETICAL

INFORMATIONAL

Benefits

Intellisense

You will love this and get addicted to his especially if you are building complex models

Multi-conditional formula

Basic ones

Sumif

Averageif

Countif

Minif

maxif

Intermediate:

These are great, but what happens when youneed more?

Medianif?

Modeif?

Percintileif

Stdevif

Geomeanif

Anyexcelformulaif

Introduce greatest formula ever - aggregate

n cell F2, type this formula =MEDIAN (IF ($A$2:$A$12=$E2,IF ($C$2:$C$12=F$1,$B$2:$B$12))), and press Shift >

Having graphs against formulas

How to reference to make it dynamic

Pivot tables

Connecting to a table

Using a pivot table basics

Pivot table naming

Forms and fields

Replicate what we did in up above

When to use pivot tables

Much easier

Much more flexibility

Only the beginning

Filter data (recreate up above)

Conditional formatting or spark cells

Intermediate skills

Change pivot table formats

View / perspective

Color formatting

Saving formatting

Reference cells (both ways, and pluses and minus to both)

Interpreting summarized data

Use statistics description - think beyond your data

Create reports summary

Limitations on multiple pivot tables

How to structure

Caution on pivot tables

Advanced

Custom formulas - I put it but only because not everyone does power bi but try not to use it

How to build reports with pivot tables or tables

Before we talk technical tips / tricks you need to understand framework

Automate, automate, automate

Variation is the enemy, consistency is the savior

Data contract with input providers

Checks to ensure you don't have things missing

Understand that we have difference between types of data

Facts vs. reference ( or dimensions)

Examples of that

Context transition

Need to fully understand how this looks in Excel vs. power pivot

If I wanted one variable to defined in context in which I created it vs the context that it currently is, how would that looks like?

Probably need to duplicate the data, eg. Have one table that has the definition context (say average of category sales) and then need another area where I execute the formula

Time Intelligence calculations in R | by Hamza Rafiq | Towards Data Science

Time intelligence functions

Library ideas

ABC analysis and variance with group vs. overall population

Nonstandard date calendars

4-5-4 calendar (according to retail federation standards for leap year treatment)

4-4-5 calendar

About

Collections of learnings of how to use R

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published