def generate_cord():
"""
TODO: Pass the parameters as input arguments.
Parameters:
- Fs : sampling frequency
- F0 : frequency of the notes forming chord
- gain : gains of individual notes in the chord
- duration : duration of the chord in second
- alpha : attenuation in KS algorithm
"""
Fs = 48000
# D2, D3, F3, G3, F4, A4, C5, G5
F0 = 440 * pylab.array(
[2**-(31.0/12), 2**-(19.0/12), 2**-(16.0/12), 2**(-14.0/12),
2**-(4.0/12), 1.0, 2**(3.0/12), 2**(10.0/12)])
gain = [1.2, 3.0, 1.0, 2.2, 1.0, 1.0, 1.0, 3.5]
duration = 4.0
alpha = 0.9785
# Number of samples in the chord.
nbsample_chord = Fs * duration
# This is used to correct alpha later, so that all the notes
# decay together (with the same decay rate).
first_duration = pylab.ceil(nbsample_chord / pylab.round_(Fs/F0[0]))
# Initialization.
chord = pylab.zeros(nbsample_chord)
for i, f in enumerate(F0):
print("Working on %g / %g" % (i+1, len(F0)))
# Get M and duration parameter.
current_M = pylab.round_(Fs/f)
current_duration = pylab.ceil(nbsample_chord / current_M)
# Correct current alpha so that all the notes decay together
# (with the same decay rate)
current_alpha = alpha ** (first_duration / current_duration)
# Let Paul's high D on the bass ring a bit longer.
if i == 1:
current_alpha = current_alpha ** 0.8
# Generate input and output of KS algorithm.
x = pylab.rand(current_M)
y = ks(x, current_alpha, int(current_duration))
y = y[:int(nbsample_chord)]
# Construct the chord by adding the generated note (with the
# appropriate gain).
chord = chord + gain[i] * y
return Fs, duration, chord
def pie_chart():
plb.figure("How do we get to work")
plb.axes([.035, 0.035, 0.9, 0.9])
l = 'car', 'truck', 'boat', 'dingie', 'train', 'plane', 'bus', 'rocket', 'tram', 'other'
b = plb.round_(plb.random(10), decimals=2)
c = ['blue', 'red', 'green', 'gray', 'yellowgreen', 'gold', 'lightskyblue', 'lightcoral', 'cyan', 'orange']
e = (0, 0, 0, 0, 0, 0, 0, 0.05, 0, 0)
plb.cla()
plb.pie(b, explode=e, labels=l, colors=c, radius=.75, autopct='%1.0f%%', shadow=True, startangle=15)
plb.axis('equal')
plb.xticks(()); plb.yticks(())
plb.pause(5)
def get_chans(msfile, fi=1418e06, di=100):
'''
For msfile, takes a desired frequency (in Hz) and returns the lower and upper bounds for a
window on either side of fi, in lengths of di
'''
tb.open(msfile+'/SPECTRAL_WINDOW');
f1 = chan_freq = tb.getcol('CHAN_FREQ')[0];
df = tb.getcol('CHAN_WIDTH')[0];
i = pl.round_((f1-fi)/df);
tb.close();
spw_lower = i - di; spw_upper = i + di;
if spw_lower < 0:
spw_lower = 0;
spw_upper = 200;
return(str(int(spw_lower)), str(int(spw_upper)));
def probeData(settings):
print "Probing data", settings.fileName
samplesPerCode = int(
round(settings.samplingFreq /
(settings.codeFreqBasis / settings.codeLength)))
samples = getSamples.int8(settings.fileName, 10 * samplesPerCode,
settings.skipNumberOfBytes)
#Initialize figure
fig = pylab.figure()
pylab.clf()
#X axis
timeScale = [x*(1/settings.samplingFreq) for x in \
range(0,int(round((5e-3 + 1/settings.samplingFreq)*settings.samplingFreq)))]
#Time domain plot
pylab.subplot(2, 2, 1)
plot_max = int(round(samplesPerCode / 50))
pylab.plot([1000 * i for i in timeScale[0:plot_max]], samples[0:plot_max])
pylab.title('Time domain plot')
pylab.xlabel('Time (ms)')
pylab.ylabel('Amplitude')
#Frequency domain plot
(Pxx,freqs) = matplotlib.mlab.psd(x = samples-numpy.mean(samples),\
noverlap = 1024,\
NFFT = 2048,\
Fs = settings.samplingFreq/1e6)
pylab.subplot(2, 2, 2)
pylab.semilogy(freqs, Pxx)
pylab.title('Frequency Domain Plot')
pylab.xlabel('Frequency (MHz)')
pylab.ylabel('Magnitude')
#Histogram
pylab.subplot(2, 2, 3)
xticks = pylab.unique(samples)
pylab.hist(samples, len(xticks))
axis = pylab.axis()
pylab.axis([min(samples), max(samples), axis[2], axis[3]])
xticks = pylab.unique(pylab.round_(xticks))
pylab.xticks(xticks)
pylab.title('Histogram')
return fig
def get_chans(msfile, fi=1418e06, di=100):
"""
For msfile, takes a desired frequency (in Hz)
and returns the lower and upper bounds for a
window on either side of fi, in lengths of di
"""
tb.open(msfile + "/SPECTRAL_WINDOW")
f1 = chan_freq = tb.getcol("CHAN_FREQ")[0]
df = tb.getcol("CHAN_WIDTH")[0]
i = pl.round_((f1 - fi) / df)
tb.close()
spw_lower = i - di
spw_upper = i + di
if spw_lower < 0:
spw_lower = 0
spw_upper = 200
return (str(int(spw_lower)), str(int(spw_upper)))
def probeData(settings):
print "Probing data", settings.fileName
samplesPerCode = int(round(settings.samplingFreq / (settings.codeFreqBasis / settings.codeLength)))
samples = getSamples.int8(settings.fileName,10*samplesPerCode,settings.skipNumberOfBytes)
#Initialize figure
fig = pylab.figure()
pylab.clf()
#X axis
timeScale = [x*(1/settings.samplingFreq) for x in \
range(0,int(round((5e-3 + 1/settings.samplingFreq)*settings.samplingFreq)))]
#Time domain plot
pylab.subplot(2,2,1)
plot_max = int(round(samplesPerCode/50))
pylab.plot([1000*i for i in timeScale[0:plot_max]],samples[0:plot_max])
pylab.title('Time domain plot')
pylab.xlabel('Time (ms)')
pylab.ylabel('Amplitude')
#Frequency domain plot
(Pxx,freqs) = matplotlib.mlab.psd(x = samples-numpy.mean(samples),\
noverlap = 1024,\
NFFT = 2048,\
Fs = settings.samplingFreq/1e6)
pylab.subplot(2,2,2)
pylab.semilogy(freqs,Pxx)
pylab.title('Frequency Domain Plot')
pylab.xlabel('Frequency (MHz)')
pylab.ylabel('Magnitude')
#Histogram
pylab.subplot(2,2,3)
xticks = pylab.unique(samples)
pylab.hist(samples,len(xticks))
axis = pylab.axis()
pylab.axis([min(samples),max(samples),axis[2],axis[3]])
xticks = pylab.unique(pylab.round_(xticks))
pylab.xticks(xticks)
pylab.title('Histogram');
return fig
def getLogCurve(group, args):
"""Return logistic curve of each Pandas GroupBy group"""
plateFlag = args[0]
finalTime = args[1][group.name] # key = (sample, well)
time = py.linspace(0.0, finalTime, 100)
logistic = GrowthCurve.logistic(time, group["y0"][0],
group["asymptote"][0],
group["maxgrowth"][0], group["lag"][0])
# Create DataFrame object with logistic curve
d = {"od": logistic}
if plateFlag:
d["mainsource"] = group["mainsource"][0]
d["compound"] = group["compound"][0]
df = pd.DataFrame(d, index=py.round_(time, decimals=3))
df.index.name = "time"
return df
def getLogCurve(group, args):
"""Return logistic curve of each Pandas GroupBy group"""
plateFlag = args[0]
finalTime = args[1][group.name] # key = (sample, well)
time = py.linspace(0.0, finalTime, 100)
logistic = GrowthCurve.logistic(time,
group["y0"][0],
group["asymptote"][0],
group["maxgrowth"][0],
group["lag"][0])
# Create DataFrame object with logistic curve
d = {"od": logistic}
if plateFlag:
d["mainsource"] = group["mainsource"][0]
d["compound"] = group["compound"][0]
df = pd.DataFrame(d, index=py.round_(time, decimals=3))
df.index.name = "time"
return df
def imslice(myimage, v, theta, x0, y0, D): ia.open(myimage); pix = pl.absolute(pl.round_((ia.summary()['incr']*180./pl.pi)*3600.)[0]) if v=='major': v = [0., D]; v2 = [0., -D]; if v=='minor': v = [D, 0]; v2 = [-D, 0]; x1 = pl.array((v)); x1r = pl.dot(rotmax(theta), x1); x2 = pl.array((v2)); x2r = pl.dot(rotmax(theta), x2); j1 = x0+x1r[0]; j2 = x0+x2r[0]; k1 = y0+x1r[1]; k2 = y0+x2r[1]; s = ia.getslice([j1, j2], [k1, k2], npts=500, method='linear') x = s['distance']*pix-D*pix; y = s['pixel']; ia.close(); return x, y
msno.matrix(train)
# ## Data Visualization
#
# ### Let's visualize the data to gain a new perspective.
#
# ### Class Distribution
#
# #### I can see from the plot below that there are more ships than icebergs in our dataset but still the class is not unbalanced.
# In[186]:
count_iceberg = train.is_iceberg.value_counts()
plt.figure(figsize=(10, 10))
b = plb.round_(count_iceberg.values)
c = ['green', 'red']
e = [0, 0.1]
l = 'Ship', 'Iceberg'
plb.pie(b,
explode=e,
labels=l,
colors=c,
radius=.9,
autopct='%1.2f%%',
shadow=True,
startangle=15)
#plt.xticks(range(2), ['Ships', 'Icebergs'])
plt.title('Class Distribution: Ship or Iceberg')
plt.show()
def print_mare(vars):
if 'p_obs' in vars:
are = pl.atleast_1d(
pl.absolute(
(vars['p_obs'].value - vars['pi'].value) / vars['pi'].value))
print 'mare:', pl.round_(pl.median(are), 2)
def curveFit(group, args):
"""Perform logistic fit on each growth curve using Pandas GroupBy groups"""
sample, rep, well = group.name
dataLogParams = args[0]
plateFlag = args[1]
growthFlag = args[2]
# Perform logistic fitting
try:
if plateFlag:
gCurve = GrowthCurve.GrowthCurve(group["od"], sample, rep, well,
growthFlag)
else:
gCurve = GrowthCurve.GrowthCurve(group, sample, rep, well,
growthFlag)
except Exception as e:
util.printStatus("sample: {}, rep: {}, well: {}".format(*group.name))
util.printStatus(e)
util.exitScript()
# Add logistic parameters to DataFrame
time = group.index.get_level_values("time")
dataLogParams["y0"].loc[sample, rep, well] = gCurve.y0
dataLogParams["maxgrowth"].loc[sample, rep, well] = gCurve.maxGrowthRate
dataLogParams["asymptote"].loc[sample, rep, well] = gCurve.asymptote
dataLogParams["lag"].loc[sample, rep, well] = gCurve.lag
dataLogParams["growthlevel"].loc[sample, rep, well] = gCurve.growthLevel
dataLogParams["glscaled"].loc[sample, rep, well] = gCurve.glScaled
dataLogParams["r"].loc[sample, rep, well] = gCurve.expGrowth
dataLogParams["auc_raw"].loc[sample, rep, well] = gCurve.auc_raw
dataLogParams["auc_rshift"].loc[sample, rep, well] = gCurve.auc_rshift
dataLogParams["auc_log"].loc[sample, rep, well] = gCurve.auc_log
dataLogParams["auc_lshift"].loc[sample, rep, well] = gCurve.auc_lshift
dataLogParams["growthclass"].loc[sample, rep, well] = gCurve.growthClass
dataLogParams["err_mse"].loc[sample, rep, well] = gCurve.mse
if plateFlag:
try:
m = group["mainsource"][0]
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to obtain 'mainsource' from "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group["mainsource"].values)
util.printStatus(group)
util.exitScript()
try:
c = group["compound"][0]
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to obtain 'compound' from "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group["compound"].values)
util.printStatus(group)
util.exitScript()
try:
dataLogParams["mainsource"].loc[sample, rep, well] = m
dataLogParams["compound"].loc[sample, rep, well] = c
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to use loc() method on "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group)
util.exitScript()
# Create Series object of logistic values to return
idxNames = ["sample", "rep", "well", "time"]
index = zip([group.name] * len(time), time)
index = [x[0] + tuple([x[1]]) for x in index]
index = pd.MultiIndex.from_tuples(index, names=idxNames)
# Create DataFrame object with new logistic curve
d = {"od": gCurve.dataLogistic}
if plateFlag:
d["mainsource"] = m
d["compound"] = c
df = pd.DataFrame(d, index=py.round_(time, decimals=3))
df.index.name = "time"
return df
plb.hist(unif_dist, bins=24, histtype='stepfilled', normed=True, color='cyan', label='Uniform')
plb.hist(gaus_dist, bins=24, histtype='stepfilled', normed=True, color='orange', label='Gaussian', alpha=0.065)
plb.legend(loc='upper left')
plb.title('Gaussian vs Uniform distribution/Histogram')
plb.xlabel('Value')
plb.ylabel('Frequency')
plb.grid(True)
plb.show()
# Pie Chart
plb.figure('How do we get to work:')
plb.axes([0.035, 0.035, 0.9, 0.9])
l = 'Car', 'Truck', 'Boat', 'Dingie', 'Train', 'Plane', 'Bus', 'Rocket', 'Tram', 'Other'
b = plb.round_(plb.random(10), decimals=2)
c = ['blue', 'red', 'green', 'gray', 'yellowgreen', 'gold', 'lightskyblue', 'lightcoral', 'cyan', 'orange']
e = (0, 0, 0, 0, 0, 0, 0, 0.05, 0, 0) #'explode' the 8th slice only
plb.cla()
plb.pie(b, explode = e, labels=l, colors=c, radius=.75, autopct='%1.2f%%', shadow=True, startangle=15)
#we set aspect ratio to 'equal' so the pie is drawn in a circle
plb.axis('equal')
plb.xticks(()); plb.yticks(())
plb.show()
# Contour plot:
def f(x,y):
return (2- x/3 + x**6 +2.125*y) * plb.exp(-x**2-y**2)
def print_mare(vars):
if 'p_obs' in vars:
are = pl.atleast_1d(pl.absolute((vars['p_obs'].value - vars['pi'].value)/vars['pi'].value))
print 'mare:', pl.round_(pl.median(are), 2)
def curveFit(group, args):
"""Perform logistic fit on each growth curve using Pandas GroupBy groups"""
sample, rep, well = group.name
dataLogParams = args[0]
plateFlag = args[1]
growthFlag = args[2]
# Perform logistic fitting
try:
if plateFlag:
gCurve = GrowthCurve.GrowthCurve(group["od"], sample, rep, well,
growthFlag)
else:
gCurve = GrowthCurve.GrowthCurve(group, sample, rep, well,
growthFlag)
except Exception as e:
util.printStatus("sample: {}, rep: {}, well: {}".format(*group.name))
util.printStatus(e)
util.exitScript()
# Add logistic parameters to DataFrame
time = group.index.get_level_values("time")
dataLogParams["y0"].loc[sample, rep, well] = gCurve.y0
dataLogParams["maxgrowth"].loc[sample, rep, well] = gCurve.maxGrowthRate
dataLogParams["asymptote"].loc[sample, rep, well] = gCurve.asymptote
dataLogParams["lag"].loc[sample, rep, well] = gCurve.lag
dataLogParams["growthlevel"].loc[sample, rep, well] = gCurve.growthLevel
dataLogParams["glscaled"].loc[sample, rep, well] = gCurve.glScaled
dataLogParams["r"].loc[sample, rep, well] = gCurve.expGrowth
dataLogParams["auc_raw"].loc[sample, rep, well] = gCurve.auc_raw
dataLogParams["auc_rshift"].loc[sample, rep, well] = gCurve.auc_rshift
dataLogParams["auc_log"].loc[sample, rep, well] = gCurve.auc_log
dataLogParams["auc_lshift"].loc[sample, rep, well] = gCurve.auc_lshift
dataLogParams["growthclass"].loc[sample, rep, well] = gCurve.growthClass
dataLogParams["err_mse"].loc[sample, rep, well] = gCurve.mse
if plateFlag:
try:
m = group["mainsource"][0]
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to obtain 'mainsource' from "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group["mainsource"].values)
util.printStatus(group)
util.exitScript()
try:
c = group["compound"][0]
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to obtain 'compound' from "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group["compound"].values)
util.printStatus(group)
util.exitScript()
try:
dataLogParams["mainsource"].loc[sample, rep, well] = m
dataLogParams["compound"].loc[sample, rep, well] = c
except TypeError as e:
stars = "*" * 55
err = "\n" + stars + "\nERROR: trying to use loc() method on "
err += " ".join((sample, rep, well)) + "\n" + stars
util.printStatus(err)
util.printStatus(e)
util.printStatus(group)
util.exitScript()
# Create Series object of logistic values to return
idxNames = ["sample", "rep", "well", "time"]
index = zip([group.name] * len(time), time)
index = [x[0] + tuple([x[1]]) for x in index]
index = pd.MultiIndex.from_tuples(index, names=idxNames)
# Create DataFrame object with new logistic curve
d = {"od": gCurve.dataLogistic}
if plateFlag:
d["mainsource"] = m
d["compound"] = c
df = pd.DataFrame(d, index=py.round_(time, decimals=3))
df.index.name = "time"
return df
