# Filter and Correlation

**Introduction :**

Filters are used to extract the details that we want in the tableau chart. Using it we can alter immediately in charts and separately view the details. If this filter card is available in the sheet it will automatically be added to the dashboards and we can filter the results in the dashboards too. Filters are used in all kinds of charts in tableau. Correlation is used to see the relation between the measures. It can be seen by including the trend lines in the graph. The region-wise sales and profit data of the superstore has been used in the below example to explain filter and correlation.

**Filter :**

As the name suggests, it filters and highlights the required data. Tableau uses it to see the precise

details of the data. Filtering is nothing but removing some scope of data from a data set. It is very helpful to create dashboards in Tableau. Filters can help to minimize the size of data sets for efficient use, eliminate irrelevant dimension elements, clean up underlying data, set date ranges and measures as required, simplify and organize data, etc. There are many types of filters available in tableau. They are

Context filter

Extract filter

Data source filter

Dimension filter

Measure filter

Table calculation filter

These are the most often used filters other than this we use a global filter,Quick filter,cascading filter, and user filter.

Let's take a region-wise profit and sales chart where profit and sales will be given in rows and columns and region is given in colors.

**Dimension filter :**

Dimension filters in Tableau are non-aggregated filters used for discrete data.

Let's see about the dimension filter in detail. we can find the filter above the marks in tableau.

Which data you want to apply filter can be dragged and dropped in the filter box. Now we want to filter the regions and see the profit and sales separately. As soon as we dropped, a checkbox will be opened as shown below.

Select the required field which wanted to be included while applying the filter. Here we want all the regions so that we can separately view all four regions and so we selected all the regions.

After applying we can notice the region in the filter. In order to get the required results, let's have the filter checkbox.

To get the filter check box click the drop-down button of the filter and select show filter.

The filter check box is visible in the right.

To extract each region separately by selecting the checkboxes. Here, we can see only the “central” and “south” region results since it has been selected in the filter option. Likewise, we can extract the required results using the filter option.

**Context Filter :**

Context filters in Tableau can help to create data sets by applying relevant presets for compilation. Tableau context filter is always processed and applicable first, even if others are applied. The multiple preset categories in the worksheet can be divided into many more parts that end up working like a context filter in itself. Data sets can be minimized efficiently by this for viewing all data rows despite the constraints. The sheets can be chosen as and when needed.

**Extract filter :**

Extract filters in Tableau are used to extract a small subset of data from the original data source. Tableau then creates a local copy of the data set that is to be stored in the repository. The data size used can be further reduced by applying the measure or dimension filter to the extract as required.

**Data source filter:**

Data source filters in Tableau are mainly used to restrict sensitive data from viewers and reduce data feeds. One important thing to mention is that the extract filter and the data source filter are not linked, and if we happen to go back to a live connection, the data source filter will remain intact.

**Measure Filter:**

Using a Measure filter in Tableau allows for various operations and aggregate functions such as sum, median, avg, standard deviation, etc. By using this we can change the ranges in the chart.

Here we have four options:

**Range:**Select the range of values to include in the result**At least:**Select the minimum value of a measure**At most:**Select the maximum value of a measure**Special:**Select null or non-null values

**Table calculation Filter :**

The Table Calculation filter is the last filter that is applied after the view has been created. If we want the calculation down the table but we have the calculation across the table at the time we can use this filter.

**Global filter :**

The Global filter can be applied across multiple worksheets by using the same source data within a workbook. The filter can be applied to all worksheets by using the same data as well.

**Quick filter:**

The various filter types in Tableau are quickly accessible by using the right-click option. These filters are known as Quick filters, and they have sufficient functionality for all common filtering needs. Quick filters in Tableau can also be implemented on dimensions or measures.

**Cascading filter:**

Cascading filters in Tableau allow for the selections in the first filter to change the options in the second filter. This helps to limit the values to ones that are only relevant to the first filter and prevents users from selecting irrelevant data, which creates a better user experience.

**User filter: **

The User filter, also known as the row-level security, in Tableau it is a feature that restricts and manages the data that users can view or access based on the authority given.

Now let's see about the correlation.

**Correlation :**

The correlation is a statistical measure of the strength of a linear relationship between two variables. It is used to see the growth relation between two factors. There are three types of correlation. These are calculated by “trend line”(A linear trend model is computed for the two given factors.).Correlation analysis can reveal meaningful relationships between different metrics or groups of metrics. Information about those connections can provide new insights and reveal interdependencies, even if the metrics come from different parts of the business. Here we used Region wise sales and profit scatter plots of the superstore.

Types of Correlation:

1. Positive correlation

2. zero correlation

3. Negative correlation

**Positive Correlation :**

When we find an inclined trend line between the axis we can say a “positive correlation” exists between those factors.

Below scatter plot is an example of a positive correlation. Here we can see the inclined trendlines between sales and profit i.e when the sales “increase” and profit “increase” simultaneously. If two-factor increases simultaneously it results in inclined trend lines in the graph then the factors are “positively correlated”.

**Zero Correlation :**

Here we can find “parallel” trend lines that mean there is “no increase” in both factors. Both will be equal for a while.

In the below graph, we can see a kind of zero relation where sales and profit are the same approximately.

**Negative Correlation:**

In this type, we can find the inverted inclined trend lines. Here one factor increases the other factor decreases.

Here we can see the trendlines coming down i.e when the sales “increase” and profit “decreases”.Negative correlation examples can be clearance sales and garage sales. when there is a loss we can find a Negative correlation.

**Benefits of Correlation :**

**Reduce Time to Detection**

In anomaly detection, working with a vast number of metrics and surfacing correlated anomalous metrics helps draw relationships that not only reduce time to detection (TTD) but also support shortened time to remediation (TTR). Anomaly detection is the identification of rare items, events, or observations that raise suspicions by differing significantly from the expected or majority of the data

##### Reduce Alert Fatigue

Another important benefit of correlation analysis in anomaly detection is in reducing alert fatigue by filtering irrelevant anomalies (based on the correlation) and grouping correlated anomalies into a single alert.

##### Reduce Costs

Correlation analysis helps significantly reduce the costs associated with the time spent investigating meaningless or duplicative alerts.

**Conclusion :**

Filters are used in all chart types. Correlation is used when we want to establish a relationship and view growth. This Correlation analysis calculates the level of change in one variable due to the change in the other. Hope this blog gives a brief about the filters and correlation. Please do follow and support for similar blogs.