Palette says it all !
On my journey of learning python and performing exploratory data analysis. I have come to realised that performing EDA and creatively visualising the data goes hand-in hand and when we say creatively visualising it includes :
- Nomenclature of Graphs
- No overlapping of attributes on the axis
- Choosing the correct colour palettes.
Somewhere themes, makes the graphs look more pleasing to the eyes and that why I believe both Matplotlib and Seaborn library has various palettes. I am going to share those which are my personal favourite which I always include while performing the EDA on any dataset.
For Charts
- palette = ‘cubehelix’
code = df[‘ ’].plot.pie(colors = sns.color_palette(‘cubehelix’)
2. palette = ‘Spectral’
Code = sns.barplot(x=‘ ’ y=‘ ’, data = , hue = ‘ ’, palette=‘Spectral’)
3. palette = ‘vlag’
Code = sns.barplot(x=‘ ’ y=‘ ’, data = , hue = ‘ ’, palette=‘vlag’)
For Heatmaps :
- cmap=‘icefire’
Code = sns.heatmap(df.corr(),annot=True, cmap=‘icefire’ )
2. cmap = ‘coolwarm’
Code = sns.color_palette(“coolwarm”, as_cmap=True)
3. cmap = ‘seagreen’
Code = sns.light_palette(“seagreen”, as_cmap=True)
Well, these are my ‘go to palette’ when it comes EDA, what are yours ?
PS — Always follow a uniform theme throughout the data set, it makes the analysis look more organised and streamlined.
Thank- you for reading it till the end. Feel free to add you suggestion.