Must-have charts for your data analysis toolbox

Tabular data has one big problem, it is incompatible with our brain. Represent the data with visuals and your analysis will shine!

The vast majority of the data is stored in tables, which are the best format for a computer to understand. However, humans are visual creatures and we are much better at understanding visual data. Representing tabular data with charts can make a big difference; nevertheless, it is important to choose the right type of chart.

Some charts are better for representing numerical data (1, 1.2, 3, 4…), others excel at visualizing categorical data (cat, dog…). Charts can also be divided by the number of data dimensions (columns of tabular data) they can visualize.

* 2 categorical and 1 numerical data dimension

Bar chart

In my opinion, it is by far the most useful chart that can be used in most situations. Bar chart is great for showing both numerical and categorical data. Usually, it represents 1 or 2 data dimensions, but you can add more using grouping (also referred to as Grouped bar chart). It can be used in the horizontal or vertical orientation (also referred to as Column chart).

Bar chart types

Line chart

Date-time data is very common in the real world and a Line chart is one of the best charts to visualize it. Line chart makes multiple data dimensions extremely easy to read, however, values have to be in a similar range. Its ability to visualize trends and seasonality is invaluable.

Line chart types


Data distribution is usually the first thing to check when dealing with numerical data and Histogram is here to do the job. It shows how often a value occurs in the dataset which enables you to detect skewness, outliers, etc. Visually it looks similar to Bar chart but they are fundamentally different.

Histogram types

Scatter plot

Scatter plot is a must-have for numerical data analysis. It is the best tool to detect the correlation between 2 columns of data, it is also great to detect outliers. Strip / Swarm plot is a special variation of Scatter plot that can represent categorical values.

Scatter plot types

Box plot

Data distribution can also be represented with various statistic metrics (median, quartiles, IQR…). Most of them are combined inside the Box plot. The chart represents those metrics in a nice compact way, which makes it perfect for comparing different distributions. However, it can be a bit unintuitive if you are not used to it.

Box plot types


Heatmap is a 3-dimensional chart that can only represent 2 categorical and 1 numerical value. It seems very specific but actually, there are many use cases for this combination of data columns. I believe Heatmap is one of the most intuitive ways to represent complex patterns with 3 data dimensions. Bubble chart can also be used to describe the same patterns.

Heatmap example
Bubble chart example

Special charts for specific data analysis:


Textual data is hard to analyze and especially visualize, however it is not impossible. Wordcloud is a great way to represent the number of distinct words/phrases in a tight space.

Wordcloud example

Radar chart

It is a unique chart for showing multi-dimensional numerical data in a way that is easy to understand.

Radar chart example


Choosing the chart is a delicate decision that can reveal or disguise information. Thus, it should be handled with care. Another important aspect is chart design, which should always be finetuned.

Data visualizations are usually the most impactful elements of your reports. Want great reports? Make great visualizations!

Leave a Comment

Your email address will not be published. Required fields are marked *