Creating a matplotlib multi-series line plot with pandas

Question

I want to create a multi series line plot showing how the occurrences of a data-frame element change over time:

I have two lists in which I have joined in a dataframe:

df = pd.DataFrame(
{'Date': datelist,
 'Category': catlist
})

I have then grouped the dataframe to show the counts of each occurrence over time:

df = df.groupby(['Date', 'Category']).size()
print df

This returns something that looks like this:

 Date        Category        
 13/02/2018  clean          2
             suspicious     1
 14/02/2018  clean          2
 19/02/2018  clean          2

I now want to create a multi-series line plot for each category type, showing how the count changes over date.

I'm really not sure how to do this using matplotlib

Joe · Accepted Answer · 2018-03-05 14:19:11Z

2

You can try with:

plot_df = df.unstack('Category')
plot_df.index = pd.PeriodIndex(plot_df.index.tolist(), freq='D')
plot_df.plot()
plt.show()

Or with subplot:

plot_df.plot(subplots=True)

Example: For the dataframe:

Date        Category
13/02/2018  clean
13/02/2018  clean
13/02/2018  suspicious
14/02/2018  clean
14/02/2018  clean
19/02/2018  clean
19/02/2018  clean
14/02/2018  suspicious
13/02/2018  suspicious
14/02/2018  clean

edited Mar 5, 2018 at 14:19

answered Mar 5, 2018 at 13:03

Joe

12.4k7 gold badges44 silver badges58 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Samolivercz Over a year ago

Thanks, is there a reason why it only plots the one category and only plots a single date from the list?

Joe Over a year ago

What do you mean? It can happen that if a category is constant, you dont see it in the graph

Samolivercz Over a year ago

My main issue now is that the Dates on the X axes Appear as the following "13 Feb 18", "14", "15" - why is the full date not shown?

Joe Over a year ago

For the suspicious category most likely is because it is constant, or just for a day. Let me check for the date

Samolivercz Over a year ago

Let us continue this discussion in chat.

devd_102 · Accepted Answer · 2018-03-05 13:05:58Z

0

categories = df.Category.unique()   #to find unique categories
f,plots = plt.subplots(1,categories.len())   #generate subplot grid (1xNo. of unique categories)
for i, category in enumerate(categories):
    temp = df['Category'==category]   #get data for each category
    plots[i].scatter(temp['Date'],temp['Count'])   #plot for that category
plt.show()   #show final generated plot

Above is python code, there might be some syntax error but it will give you an idea how to solve your problem and how to code.

answered Mar 5, 2018 at 13:05

devd_102

235 bronze badges

Collectives™ on Stack Overflow

Creating a matplotlib multi-series line plot with pandas

2 Answers 2

5 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

5 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related