stripplotfromcsv.py

#making a strip plot from a csv file using seaborn
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import csv
import sys

filename = sys.argv[1]
reader = csv.DictReader(open(filename))

my_data = {}
for row in reader:
    for column, value in row.iteritems():
        my_data.setdefault(column, []).append(value)
print my_data

# Format into dictionary
methods_list = []
error_list = []


#for key in my_data: #for some reason the order is screwed up using this option
for key in ["NULL", "PM6-DH+/C", "PM6/C", "PM3/C", "AM1/C", "PM3/S", "AM1/S", "DFTB3/S"] :
    for data in my_data[key]:
        methods_list.append(key)
        error_list.append(data)

method_error = dict()
method_error["method"] = methods_list
method_error["error"] = error_list

# Save as panda data frame
my_data_formatted = pd.DataFrame(method_error)

# Print to see format
#print my_data_formatted

sns.stripplot(x="method", y="error", data=my_data_formatted, jitter=True,
              palette="Set2",split=True,linewidth=1,edgecolor='gray')

out = filename[:-4]+".png"
plt.savefig(out)