Matplotlib. Plotting in 'for loops' to same axis

2
votes

I have a rather simple problem but something that's had me stumped for 2days. I need to plot 2+ files. Each file will need to be plotted on a total of 25 plots but must be plotted on the same set of axes. (ie. if theres 2 files I need 25 plots with 2 lines on each plot).

I have this sudo code which generates 50 plots (One line for each)...which is wrong

with open(bamlist, 'r') as bamlist:
    for bam in bamlist:     #Opens the 2 files
        'Generate data Here'
        dataframe = []
        for line in data:
            line = line.split("\t")
            dataframe.append(line[0:4:1])
        df = pd.DataFrame(dataframe, columns=['Chromosome', 'Position', 'N', 'BaseCount'])

        grouped_df = df.groupby('Chromosome')     #groups dataframe into the required 25plots
        for df in grouped_df:
            density_data = 'Get density data from df'
            f, ax = plt.subplots()          
            sns.kdeplot(density_data, ax=ax, linewidth=1)
            pp.savefig()
pp.close()

Is there a way to revert back to the initial set of axis the 2nd time the for loop is entered so that I will get 2 lines per plot with 25plots (as opposed to 50)?

2
pythonmatplotlib
do you want 25 subplots on one figure, or 25 separate figures? - tmdavison
Either will do. separate figures will be easier to view tho. - Michael Berry

2 Answers

8
votes

Your problem stems from your use of:

f, ax = plt.subplots()

This means that you generate a new subplot every time you hit that line (in your case, 50 times). You need to generate 25 subplots and reference them later on. You can do something like:

axes = []
for i in range(25):
    f,ax = plt.subplots()
    axes.append(ax)

Then in your loop:

for df_index in range(len(grouped_df)):
    df = grouped_df[df_index]
    density_data = 'Get density data from df'
    sns.kdeplot(density_data, ax=axes[df_index], linewidth=1)

You can also do a check to see if the axis doesn't exist (if it extends to more than 25 subplots or something), and if not, create it.

3
votes

Using plt.figure() also gets the job done...

iterate = -1
for df in grouped_df:
    iterate += 1
    plt.figure(iterate)
    density_data = 'Get density data from df'
    sns.kdeplot(density_data, linewidth=1)
    pp.savefig()
pp.close()