lol learningcurve

plot.py

#!/usr/bin/env python2

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
import matplotlib

matplotlib.rcParams['xtick.minor.size'] = 0
matplotlib.rcParams['xtick.minor.width'] = 0

cmat_H = np.array([0.01411625284730, 0.01156763771220, 0.01028136505150, 0.00922856778135,
                   0.00832607218566, 0.00753118369583, 0.00680493804346])

cmat_C = np.array([0.05209449056730, 0.04311612940300, 0.03621917068040, 0.03045724774640,
                   0.02567189952020, 0.02190313628230, 0.01860507453570])
    
cmat_N= np.array([0.0594763460734,0.0498454460526,0.0418617840876,0.0344605890075,
                  0.0287360921768,0.0235323935813,0.0188990017579])



cmat_O = np.array([ 0.0351159819664, 0.0306679911167, 0.0261231594199, 0.0219098131318,
                    0.0183183159273, 0.0148631312546, 0.0120398243948])



arad_H = np.array([0.00524686981787, 0.00473619017511, 0.00444523921688, 0.00420098359973,
                   0.00404343236846, 0.00390155524685, 0.00377452758820])

arad_C = np.array([
    #0.0187120144857,
    #0.0153668990028,
    #0.0134273947451,
    #0.0123309197662,
    #0.0114507572425,
    #0.0108264804567])
0.0187820298058, 
0.015345503611  ,
0.0132067065705 ,
0.01205056773  ,
0.0110997907916 ,
0.0103383808794 ,
0.00977683984923])

arad_N = np.array([
    0.0164562555084 ,
    0.0140631254927 ,
    0.0123373305615 ,
    0.0111841105441 ,
    0.0104071189521 ,
    0.00974294172462,
    0.0091474878246 ])

arad_O = np.array([
0.00905516433996,
0.00766256408912,
0.00668741322232,
0.0060874760913 ,
0.0056748987601 ,
0.00538127070123,
0.00513132918469])









def major_formatter(x, pos):
    return "%2.0f" % x

def minor_formatter(x, pos):
    return "%4.3f" % x

def off_formatter(x, pos):
    return ""

sizes = np.array([1000, 2000, 4000, 8000, 16000, 32000, 64000])

ax = plt.subplot(1, 1, 1)

plt.loglog(sizes, cmat_H, color="gray",  linestyle="--", marker=".", label="H/CM")
plt.loglog(sizes, cmat_C, color="black", linestyle="--", marker=".", label="C/CM")
plt.loglog(sizes, cmat_N, color="blue",   linestyle="--", marker=".", label="N/CM")
plt.loglog(sizes, cmat_O, color="red",  linestyle="--", marker=".", label="O/CM")

plt.loglog(sizes, arad_H, color="gray",  marker=".", label="H/ARAD")
plt.loglog(sizes, arad_C, color="black", marker=".", label="C/ARAD")
plt.loglog(sizes, arad_N, color="blue",   marker=".", label="N/ARAD")
plt.loglog(sizes, arad_O, color="red",  marker=".", label="O/ARAD")

ax.set_xticks(sizes)
ax.xaxis.set_major_formatter(ticker.FuncFormatter(major_formatter))
ax.xaxis.set_minor_formatter(ticker.FuncFormatter(off_formatter))

# ax.set_yticks(np.array([0.003, 0.004, 0.005, 0.006, 0.007, 0.008,
#                         0.009, 0.010, 0.011, 0.012, 0.013, 0.014, 0.015]))
ax.set_yticks(np.array([0.003, 0.004, 0.005, 0.006, 0.007, 0.008,
                        0.009, 0.010, 0.020, 0.030, 0.040, 0.050, 0.060]))

ax.yaxis.set_major_formatter(ticker.FuncFormatter(minor_formatter))
ax.yaxis.set_minor_formatter(ticker.FuncFormatter(off_formatter))

ax.set(ylabel='MAE partial charge [e]',
       xlabel='# Atoms (training)',
       title="DFTB3 Mulliken charge prediction // GDB9")

ax.set_xlim(1000, 64000)
ax.set_ylim(0.003, 0.06)

plt.legend(loc="upper right", ncol=2)
plt.grid(True)

plt.savefig("mae.png")