In the demo to run python on the open science grid (OSG), we generated compressed numpy files that contain our results.
In this notebook, we extract these results and plot them on our data to confirm that our program ran correctly.
In this version, I am running it from my Mac. I made and used a 'python27' virt env in anaconda and ran python fabfile_getdataonly.py
In [2]:
import numpy as np
Nfiles = 10
Ps = np.zeros(Nfiles, dtype=np.ndarray)
Ts = np.zeros(Nfiles, dtype=np.ndarray)
for n in range(Nfiles):
Ps[n] = np.load('/gh/data/example/lfp_set_PsTs2/' + str(n) + '/out/Ps_data.npy')
Ts[n] = np.load('/gh/data/example/lfp_set_PsTs2/' + str(n) + '/out/Ts_data.npy')
In [3]:
lfps = np.zeros(Nfiles, dtype=np.ndarray)
for n in range(Nfiles):
if n == 0:
lfps[n] = np.load('/gh/data/example/lfp_set/' + str(10) + '.npy')
else:
lfps[n] = np.load('/gh/data/example/lfp_set/' + str(n) + '.npy')
In [4]:
import matplotlib.pyplot as plt
%matplotlib inline
plt.figure(figsize=(10,10))
for n in range(Nfiles):
plt.subplot(Nfiles, 1, n+1)
plt.plot(lfps[n], 'k')
plt.plot(Ps[n], lfps[n][Ps[n]], 'bo')
plt.plot(Ts[n], lfps[n][Ts[n]], 'ro')
if n == Nfiles-1:
plt.xlabel('Time (ms)')
else:
plt.xticks([])
plt.ylim((-3000,3000))
plt.yticks([-3000,0,3000])
if n == 0:
plt.ylabel('Voltage (uV)')
In [ ]: