def errorplot(dir_name, ax=None, fits=[1.4, 0.9]):
file_list = glob.glob(dir_name + "/Hop*.txt")
file_listN = glob.glob(dir_name + "/Notfield*.txt")
file_list = np.sort(file_list)
file_listN = np.sort(file_listN)
numfiles = len(file_list)
numfilesN = len(file_listN)
Delta = np.zeros(numfiles)
DeltaN = np.zeros(numfilesN)
Error = np.zeros(numfiles)
Errorb = np.zeros(numfiles)
for i in range(numfiles):
Delta[i] = get_pars(file_list[i])
data = np.loadtxt(file_list[i], skiprows=10).T
N = data[[1, 2, 3, 4], 0].sum()
idx = int(.25 * data.shape[1])
idxb = int(.75 * data.shape[1])
Error[i] = data[12, idx:].mean()
Errorb[i] = data[12, idxb:].mean()
ErrorN = np.zeros(numfilesN)
ErrorNb = np.zeros(numfilesN)
for j in range(numfilesN):
DeltaN[j] = get_pars(file_listN[j])
dataN = np.loadtxt(file_listN[j], skiprows=10).T
Nn = dataN[[1, 2, 3, 4], 0].sum()
idxN = int(.25 * dataN.shape[1])
idxbN = int(.75 * dataN.shape[1])
ErrorN[j] = dataN[12, idxN:].mean()
ErrorNb[j] = dataN[12, idxbN:].mean()
srtidx = np.argsort(Delta)
srtidxb = np.argsort(Delta)
srtidxN = np.argsort(DeltaN)
srtidxbN = np.argsort(DeltaN)
Delta = Delta[srtidx]
Error = Error[srtidx]
Errorb = Errorb[srtidxb]
DeltaN = DeltaN[srtidxN]
ErrorN = ErrorN[srtidxN]
ErrorNb = ErrorNb[srtidxbN]
if ax == None:
ax = plt.gca()
ax.set_title("I./C.")
ax.set_xlabel("$\Delta$")
ax.set_ylabel("Hopfield's Error Rate versus $\Delta$")
hlabel = "Hopfield"
nlabel = "Notfield"
fsl = 12
else:
ax.set_ylabel("Error Rate Ratio")
hlabel = None
nlabel = None
fsl = 12
# plt.plot(Delta, Error, "rx-", markersize = 10, label = "0.25 Data Hopfield")
ax.plot(Delta, Errorb, "bo", label=hlabel)
ax.plot(Delta,
Delta**(-fits[0]),
"b:",
label="$\Delta^{-" + str(fits[0]) + "}$")
ax.plot(Delta, Delta**(-2), "b--", label="$\Delta^{-2}$")
# plt.plot(DeltaN, ErrorN, "bx-", markersize = 10, label = "0.25 Data Notfield")
ax.plot(DeltaN, ErrorNb, "ro", label=nlabel)
ax.plot(Delta,
Delta**(-fits[1]),
"r:",
label="$\Delta^{-" + str(fits[1]) + "}$")
ax.plot(Delta, Delta**(-1), "r--", label="$\Delta^{-1}$")
ax.set_yscale('log')
ax.grid(True)
ax.legend(loc="lower left", fontsize=fsl)
return