An analysis is the foundation for visualizing your datasets. It is a container for a set of related visuals. Each analysis also contains a default story, which can be used to step through different iterations of an analysis over time; for example, you may want to share the same visual but with different filters like year/region and see them in a sequence as scenes.
An analysis is generally based on one dataset but QuickSight allows you to join datasets under one analysis, although a given visual can only have data from one dataset. You can have a maximum of 20 visuals in an analysis.
Let's look into the detailed steps for creating a new dataset and then an analysis on top of that dataset.
Before we can create an analysis, we need a dataset. Let's review the steps to create a new dataset:
- For this example, we will create a new dataset based on a USA census data file that we have in S3 storage called
USACensusSalarybyZip.csv. Click first on Manage data and then on New data set as shown in the following screenshot:
Figure 4.2: Creating a new dataset
Note
The S3 data file
USACensusSalarybyZip.csvis at the following GitHub location: https://github.com/rnadipalli/quicksight/blob/master/sampledata/USACensusSalarybyZip.csv. - Enter the Data source name, upload the S3 manifest file and then click Connect as shown in the following screenshot:
Figure 4.3: Uploading S3 data
Note
For your reference, the manifest file for the S3 data file is in the following GitHub location: https://github.com/rnadipalli/quicksight/blob/master/sampledata/USACensus-manifest.json.
- Now the source data is imported to a dataset in QuickSight. This is shown in the next screenshot. In this page, provide a name for the dataset; add column headers by double-clicking in the gray column headers and finally save. This will create a new dataset with the name,
USACensus-S3-DataSet.
Figure 4.4: Save dataset
This completes the creation of the dataset of USACensus-S3-DataSet which will appear as one of your datasets under Manage data.
Now that we have the dataset, let's build a new analysis using that dataset, which will act as the container for visuals. Here are the steps to create a new analysis:
- From the QuickSight home page, click on New analysis. This will show the list of existing datasets. Select the newly created dataset
USACensus-S3-DataSetas shown in the following screenshot:
Figure 4.5: Create analysis - use existing dataset
- After you select your dataset, you will see a pop-up as shown in the following screenshot; click on Create Analysis to use that dataset for your analysis.
Figure 4.6: Create analysis pop-up
- You will see the analysis page show up where you can change the analysis name (optional) in the application bar.
This completes the creation of a new data analysis with the name USACensus-S3-DataSet analysis. Next, we will add a visual to the analysis.
Now that we have an analysis (container), we can add visuals to the analysis using the following steps:
- In the application bar, click on Add Visual.
- Next, in the Fields list, configure your dimensions and measures. Dimensions are typically lists of values like products. Customers and measures are metrics like customer score and product quality. In our example, set the
ZipCodeas dimension and leave theMeanIncome,MedianIncomeandPopulationCountas measure. To change a field to a measure, click on the v icon next to the field of interest. - Choose the appropriate visual. For our example, let's select bar chart; see the following screenshot:
Figure 4.7: Select visualization type, dimension, and measures
- In our example, the Y axis median income should be average and this can be changed by clicking on the > sign next to the MedianIncome(SUM) text as shown in the following screenshot:
Figure 4.8: Change Y axis aggregation type
- Let's do advance formatting to the
MedianIncomefield by first selecting the field, and then click on More formatting options as shown in the following screenshot:
Figure 4. 9: Median income formatting
- Next from the Format pane on the left, select the Data Type as Currency, Decimal places as
0and Units as Thousands. This will make the Y axis easy to read.
Figure 4.10: Median income advanced formatting options
- Next, for the finishing touch, change the
ZipCodeformatting to a standard number with no comma separator. Now the chart is finally ready for presentation and is shown in the next screenshot:
Figure 4.11: Median income by zip code final
This completes the creation of a new data analysis with simple bar chart visualization.
We can rename an existing analysis by first selecting the desired analysis from the QuickSight home page and then editing the text next to analysis name. We can also add title and add description by clicking on the + add button to the left of the analysis name as shown in the following screenshot:
Figure 4.12: Rename analysis
This concludes the section for editing an existing analysis. Next let's check out how to delete an analysis.
We can delete an existing analysis by first selecting the desired analysis from the QuickSight home page and then clicking on Delete as shown in the next screenshot.
You will be prompted for a confirmation. Click on Delete to complete the deletion.
Figure 4.13: Deleting analysis
This concludes the section Building analyses from datasets; in the next section, we will review the best practices for selection of a chart type.