import matplotlib.pyplot as plt
import seaborn as sns
[docs]
def main():
sns.set()
smooth = False
# data_path_jit = f'/home/dyuman/Documents/Mancalog/profiling/profile_jit.csv'
# data_path_old = f'/home/dyuman/Documents/Mancalog/profiling/profile_old.csv'
# data_path_oldest = f'/home/dyuman/Documents/Mancalog/profiling/profile_oldest.csv'
# data = f'/home/dyuman/Downloads/memory.csv'
x_axis_title = 'Number of Nodes (density=4.10e-4)'
y_axis_title = 'Runtime (s)'
title = 'Number of Nodes vs Runtime'
# headers = ['Timesteps', 'Memory', 'Memory Old']
# df_jit = pd.read_csv(data_path_jit, names=headers, header=None)
# df_old = pd.read_csv(data_path_old, names=headers, header=None)
# df_oldest = pd.read_csv(data_path_oldest, names=headers, header=None)
# df = pd.read_csv(data, names=headers, header=None)
# x_jit = df_jit['Timesteps']
# y_jit = df_jit['Time']
# x_old = df_old['Timesteps']
# y_old = df_old['Time']
# x_oldest = df_oldest['Timesteps']
# y_oldest = df_oldest['Time']
# x = df['Timesteps']
# y1 = df['Memory']
# y2 = df['Memory Old']
x = [1000, 2000, 5000, 10000]
y2 = [0.35, 0.40, 1.35, 4.25]
y5 = [0.43, 0.49, 1.76, 6.11]
y15 = [0.40, 0.73, 3.09, 11.20]
# if smooth:
# y_jit = pd.Series(y_jit).rolling(15, min_periods=1).mean()
# y_old = pd.Series(y_old).rolling(15, min_periods=1).mean()
# y_oldest = pd.Series(y_oldest).rolling(15, min_periods=1).mean()
# sns.relplot(data=df, x =x_axis_title, y=y_axis_title, kind = 'line', hue = 'type', palette = ['red', 'blue'])
# ax = sns.relplot(data=df, kind = 'line', ci=0)
# plt.plot(x_jit, y_jit, label='accelerated with numba for CPU')
# plt.plot(x_old, y_old, label='first optimized version')
# plt.plot(x_oldest, y_oldest, label='original version')
plt.plot(x, y2, linestyle='dotted', marker='^', label='2 Timesteps')
plt.plot(x, y5, linestyle='dotted', marker='s', label='5 Timesteps')
plt.plot(x, y15, linestyle='dotted', marker='o', label='15 Timesteps')
plt.legend()
plt.title(title, fontsize=20)
plt.xlabel(x_axis_title, fontsize=13)
plt.ylabel(y_axis_title, fontsize=13)
# ax = sns.lineplot(x=x, y=y, ci=95)
# ax = sns.lineplot(x=x_jit, y=y_jit, label='')
# ax = sns.lineplot(x=x, y=y, label='')
# ax = sns.lineplot(x=x, y=y, label='')
# ax.axes.set_title(title, fontsize=18)
# ax.set_xlabel(x_axis_title, fontsize=13)
# ax.set_ylabel(y_axis_title, fontsize=13)
# plt.show()
if smooth:
plt.savefig('timesteps_vs_time_smooth.png')
else:
plt.savefig('timesteps_vs_memory.png')
if __name__ == '__main__':
main()