Subplots of bar chart for each row in pandas dataframe- Invalid color error-CodePudding

I'm creating multiple subplots (bar charts) from pandas dataframe. I would like the two bars in each plot to have different colours (so that one is always eg. orange, and the other always blue).

my bar charts

My original dataframe looks like that:

   Triplet sequence          3          4
0               AAA  78.415603  98.189959
1               AAC  85.606777  98.792146
2               AAG  85.229351  97.165199
3               AAT  78.397072  99.430332
4               ACA  85.413897  98.212358
..              ...        ...        ...
59              TGT  85.516804  98.094537
60              TTA  78.197088  99.040233
61              TTC  85.347600  98.243830
62              TTG  85.546867  98.017210
63              TTT  78.257015  99.585926

Then, I'm using a transpose to get this:

df2 = df.set_index('Triplet sequence').T
[64 rows x 3 columns]
Triplet sequence        AAA        AAC        AAG        AAT        ACA  \
3                 78.415603  85.606777  85.229351  78.397072  85.413897   
4                 98.189959  98.792146  97.165199  99.430332  98.212358

Eventually, my code for barcharts is:

bar = df2.plot(kind='bar' ,
        stacked=False,figsize = (20,40),
        title='Grouped Bar Graph with dataframe',
         subplots = True, layout=(16,4), legend = False, sharey = True ,
         fontsize = 5, ylim = (95,100), color=[['C0','C1']] )

Then I get an error: Invalid color ['C0', 'C1']. How can I specify color parameter so I get different color for each bar representing a row in my dataframe?

Thank you!

CodePudding user response：

I took the data you posted and updated it a little so that there are some rows that have both columns with values > 95. My plotted data looks like this.

df2 = df.set_index('Triplet sequence').T
df2
>>
Triplet sequence    AAA AAC AAG AAT ACA TGT TTA TTC TTG TTT
3   98.415603   85.606777   85.229351   98.397072   96.413897   85.516804   98.197088   85.34760    99.546867   78.257015
4   98.189959   98.792146   79.165199   99.430332   98.212358   98.094537   79.040233   98.24383    98.017210   99.585926

Adding colors within df2.plt() may not be the way to get each of the bars in different colors. You would need to update the bars in each graph to show the right color sequence. Your code has been modified and same is below...

bar = df2.plot(kind='bar' ,
        stacked=False,figsize = (20,20),
        title='Grouped Bar Graph with dataframe',
         subplots = True, layout=(5,4), legend = False, sharey = True ,
         fontsize = 5, ylim = (95,100))

for graph_row in bar: # Gets you a row of subplots
    for grph in graph_row : #Gets you the graph
        for ptch, color in zip(grph.patches, ["tab:orange", "tab:blue"]): #The bar
            ptch.set_facecolor(color)

will you you the below bar. Note that you can change the tab:orange and tab:blue to any other color. Also, I am assuming that each graph has just two bars. If 3, the sequence will be orange-blue-orange and so on.

The output graph is as below.