Unit 5

Prepared by: Varun Rao (Dean, Data Science & AI)

For: Data Science - 2nd years

Generating new data with forecasts

To create a forecast, your view must be using at least one date dimension and one
measure.

To turn forecasting on, right-click on the visualization and choose Forecast >Show
Forecast, or choose Analysis >Forecast >Show Forecast.

A time series is a sequence of observations recorded over a certain period of

time. A simple example of time series is how we come across different
temperature changes day by day or in a month.

Basics of Time-Series Forecasting

Time Series forecasting in simple words means to forecast or to predict the

future value(eg-stock price) over a period of time. There are different methods

to forecast the values.

while Forecasting time series values, 3 important terms need to be taken care

of and the main task of time series forecasting is to forecast these three

terms.

1) Seasonality
Seasonality is a simple term that means while predicting a time series data

there are some months in a particular domain where the output value is at a

peak as compared to other months. for example, if you observe the data of

tours and travel companies of past 3 years then you can see that in November
and December the distribution will be very high due to the holiday season and

festival season. So while forecasting time series data we need to capture this

seasonality.

2) Trend
The trend is also one of the important factors which describe that there is

certainly an increasing or decreasing trend time series, which actually means

the value of organization or sales over a period of time and seasonality is

increasing or decreasing.

3) Unexpected Events
Unexpected events mean some dynamic changes occur in an organization, or

in the market which cannot be captured. For example, the current pandemic

we are suffering from, and if you observe the Sensex of the nifty chart there is

a huge decrease in stock price which is an unexpected event that occurs in

the surrounding area.

Methods and algorithms are using which we can capture seasonality and

trend But the unexpected event occurs dynamically so capturing this becomes

very difficult.

Each of the following examples indicate the structure that supports creating a forecast.

● The field you want to forecast is on the Rows shelf and a continuous date field is
on the Columns shelf.
 ● The field you want to forecast is on the Columns shelf and a continuous date
field is on the Rows shelf.
● The field you want to forecast on either the Rows or Columns shelf, and discrete
dates are on either the Rows or Columns shelf. At least one of the included date
levels must be Year.
● The field you want to forecast is on the Marks card, and a continuous date or
discrete date set is on Rows, Columns or Marks.

With forecasting on, Tableau visualizes estimated future values of the measure, in
additional to actual historical values. The estimated values are shown by default in a
lighter shade of the color used for the historical data:

Prediction Intervals
The shaded area in the image above shows the 95% prediction interval for the forecast.
That is, the model has determined that there is a 95% likelihood that the value of sales
will be within the shaded area for the forecast period. You can configure the confidence
level percentile for the prediction bands, and whether prediction bands are included in
the forecast, using the Show prediction intervals setting in the Forecast Options
dialog box:
Clear the check box if you do not want to display prediction bands in forecasts. To set
the prediction interval, select one of the values or enter a custom value. The lower the
percentile you set for the confidence level, the narrower the prediction bands will be.

How your prediction intervals are displayed depends on the mark type of your
forecasted marks:

Forecast mark type Prediction intervals displayed using

Line Bands

Shape, square, circle, bar, or pie Whiskers

In the following example, forecast data is indicated by lighter shaded circles, and the
prediction intervals are indicated by lines ending in whiskers:

Steps in Time Series Analysis

A forecasting task usually involves five basic steps.

Step 1: Problem definition.

Often this is the most difficult part of forecasting. Defining the problem

carefully requires an understanding of the way the forecasts will be used, who

requires the forecasts, and how the forecasting function fits within the

organisation requiring the forecasts. A forecaster needs to spend time talking

to everyone who will be involved in collecting data, maintaining databases, and

using the forecasts for future planning.

Step 2: Gathering information.

There are always at least two kinds of information required: (a) statistical data,

and (b) the accumulated expertise of the people who collect the data and use

the forecasts. Often, it will be difficult to obtain enough historical data to be

able to fit a good statistical model. However, remember that good statistical

models will handle evolutionary changes in the system; don’t throw away good

data unnecessarily.

Step 3: Preliminary (exploratory) analysis.

Always start by graphing the data. Are there consistent patterns? Is there a

significant trend? Is seasonality important? Is there evidence of the presence

of business cycles? Are there any outliers in the data that need to be explained

by those with expert knowledge? How strong are the relationships among the

variables available for analysis? Various tools have been developed to help with

this analysis.
Step 4: Choosing and fitting models.

The best model to use depends on the availability of historical data, the

strength of relationships between the forecast variable and any explanatory

variables, and the way in which the forecasts are to be used. It is common to

compare two or three potential models. Each model is itself an artificial

construct that is based on a set of assumptions (explicit and implicit) and

usually involves one or more parameters which must be estimated using the

known historical data.

Step 5: Using and evaluating a forecasting model.

Once a model has been selected and its parameters estimated, the model is

used to make forecasts. The performance of the model can only be properly

evaluated after the data for the forecast period have become available. A

number of methods have been developed to help in assessing the accuracy of

forecasts. There are also organizational issues in using and acting on the

forecasts. When using a forecasting model in practice, numerous practical

issues arise such as how to handle missing values and outliers, or how to deal

with short time series.

Providing self evidence adhoc analysis with
parameters
Ad-hoc analysis is any business report or data analysis curated and created by users, as
and when they need it. Ad-hoc analysis in business intelligence sits in contrast with the
managed reports seen in the early days of business analytics, which is based on
information distributed by IT departments.

An ad-hoc analysis lets the user determine which data sources to pull from and how that
data will be presented. One key distinguishing factor of ad-hoc analysis is its ability to offer
completely customized analytics.

Parameters allow those consuming reports to change the context of views with Quick-Filter-
like controls. Report builders design parameters into views when the report is created in
Tableau Desktop. Parameters create a pathway for non-technical consumers to conduct ad
hoc analysis by changing what and how facts and dimensions are displayed—within the
boundaries of the designer’s intended usage. Concerns regarding the efficacy of self -
service analysis are minimized because the report designer controls what changes are
permitted.

Three ways Tableau facilitates ad hoc analysis are:

● Generating new data with forecasts

● Designing flexible views using parameters
● Changing or creating designs in Tableau Server

What Are Parameters?

Parameters are useful when you want to add interactivity and flexibility to your report, or to
experiment with what-if scenarios.

Parameters are dynamic values that can replace constant values in calculations, filters, and
reference lines. For example, you may create a calculated field that returns true if Sales is
greater than $500,000 and otherwise return false. You can replace the constant value of
“500000” in the formula with a parameter.
It allows users to alter the content of the formula or change a dimension or measure
contained in the view. Parameters create a powerful means for changing normally static
values into dynamic entities that facilitate ad hoc discovery without the need for changing
the design of the view.

How Can Parameters Be Used?

When you work with parameters, consider the following two things that are important in
making them useful:

● They need to be used in calculations.

● The parameter control needs to be displayed so that viewers can interact with it.

The different ways parameters can be used is limited only by your imagination. Tableau
provides some basic parameter controls by building them in different contexts that
commonly benefit from the use of a variable. Creative report designers can dream up a
myriad of other ways to use this powerful feature by building formula variables that control
the facts in view, the dimensions that appear, or the length and granularity of time-series
data. Anywhere that you can place a field in Tableau Desktop is a potential repository for
parameter control.

● In Calculations: Parameters give you a way to dynamically modify values in a

calculation. Rather than manually editing the calculation (and all dependent
calculations), you can use a parameter. Then when you want to change the value,
you open the parameter control, change the value, and all of the calculations that
use that parameter are updated.
● In Filters: Parameters give you a way to dynamically modify values in a TopN filter.
Rather than manually setting the number of values you want to show in the filter, you
can use a parameter. Then when you want to change the value, you open the
parameter control and the filter updates. For example, when creating a filter to show
the Top 10 products based on total profit, you may want to use a parameter instead
of the fixed “10” value. That way, you can then quickly update the filter to show the
top 10, 20, or 30 products.
● In Reference Lines: Parameters give you a way to dynamically modify a reference
line. For example, instead of showing a reference line at a fixed location on the axis,
you can reference a parameter. Then you can use the parameter control to move the
reference line.

Editing views in tableau server

If you can see the Edit button when you are looking at a view in Tableau Online or
Tableau Server, that means you can make changes to it. Depending on your access
level and permissions, you can:
 ● Edit an existing published workbook and add worksheets for views, dashboards,
and stories.
● Create and edit a new workbook based on a published data source.
● Edit an existing workbook and add worksheets in the browser, or by opening the
workbook in Tableau Desktop.
● Connect to different published data sources while editing.

Edit a published view

1. Sign in to a site, then either open the workbook that contains the view you want to edit,
or show All Views from the Explore page.
2. Open the view.
3. Click Edit in the view toolbar.

When you click Edit, the view opens in web authoring mode.

Now you can edit the view.

The Save options available to you will vary depending on your permissions set by your
Tableau site administrator.
● To close the view without saving your changes, select File > Close.
● To save your work in the current workbook, select File > Save.
● To save your work as a new workbook, select File > Save As.