How can the optimality gap (relative and absolute) be obtained using CPLEX Python API (CPLEX 12.5, Python 2.7.15)? Is there any function such as get_optimal_gap() that would give the optimality gap? Or is parsing the output (as mentioned here) the only way? I see that there are custom functions such as solution.get_objective_value() - it would be nice if someone can suggest online resources that have a list of all functions that can be applied on the cplex object/file. Currently I am using the following code (courtesy: this IBM document):
import cplex
import sys
def sample1(filename):
c = cplex.Cplex(filename)
try:
c.solve()
except CplexSolverError:
print "Exception raised during solve"
return
# solution.get_status() returns an integer code
status = c.solution.get_status()
print "Solution status = " , status, ":",
print c.solution.status[status]
print "Objective value = " , c.solution.get_objective_value()
sample1(filename)
The documentation for the CPLEX Python API is here (currently version 12.8).
To get the relative mip gap, you can use c.solution.MIP.get_mip_relative_gap(). To calculate the absolute mip gap you'll need c.solution.MIP.get_best_objective().
If you haven't already, you'll also want to take a look at the CPX_PARAM_EPGAP and CPX_PARAM_EPAGAP parameters.
Related
When using pvlib (but also the spectrl2 implementation provided by NREL), I obtain negative Irradiance for a north-facing panel.
Is this expected behaviour? Should the spectrum simply be cut at zero?
Added example code based on the tutorial below:
## Using PV Lib
from pvlib import spectrum, solarposition, irradiance, atmosphere
import pandas as pd
import matplotlib.pyplot as plt
# assumptions from the technical report:
lat = 49.88
lon = 8.63
tilt = 45
azimuth = 0 # North = 0
pressure = 101300 # sea level, roughly
water_vapor_content = 0.5 # cm
tau500 = 0.1
ozone = 0.31 # atm-cm
albedo = 0.2
times = pd.date_range('2021-11-30 8:00', freq='h', periods=6, tz="Europe/Berlin") # , tz='Etc/GMT+9'
solpos = solarposition.get_solarposition(times, lat, lon)
aoi = irradiance.aoi(tilt, azimuth, solpos.apparent_zenith, solpos.azimuth)
# The technical report uses the 'kasten1966' airmass model, but later
# versions of SPECTRL2 use 'kastenyoung1989'. Here we use 'kasten1966'
# for consistency with the technical report.
relative_airmass = atmosphere.get_relative_airmass(solpos.apparent_zenith,
model='kasten1966')
spectra = spectrum.spectrl2(
apparent_zenith=solpos.apparent_zenith,
aoi=aoi,
surface_tilt=tilt,
ground_albedo=albedo,
surface_pressure=pressure,
relative_airmass=relative_airmass,
precipitable_water=water_vapor_content,
ozone=ozone,
aerosol_turbidity_500nm=tau500,
)
plt.figure()
plt.plot(spectra['wavelength'], spectra['poa_global'])
plt.xlim(200, 2700)
# plt.ylim(0, 1.8)
plt.title(r"2021-11-30, Darmstadt, $\tau=0.1$, Wv=0.5 cm")
plt.ylabel(r"Irradiance ($W m^{-2} nm^{-1}$)")
plt.xlabel(r"Wavelength ($nm$)")
time_labels = times.strftime("%H:%M %p")
labels = [
"AM {:0.02f}, Z{:0.02f}, {}".format(*vals)
for vals in zip(relative_airmass, solpos.apparent_zenith, time_labels)
]
plt.legend(labels)
plt.show()
No, this is not expected behavior. I suspect the issue is caused by improper handling of angle-of-incidence values greater than 90 degrees, and essentially the same problem (for a different function) discussed here: https://github.com/pvlib/pvlib-python/issues/526
It's unfortunate that the reference implementation from NREL has the problem too (perhaps when the model was originally designed, nobody could conceive of a panel facing away from the sun!), but I think the pvlib implementation should be fixed regardless. I encourage you to file a bug report here: https://github.com/pvlib/pvlib-python/issues
In the meantime, I think you can resolve the issue in your own code by adding a line like aoi[aoi > 90] = 90 prior to passing it to spectrum.spectrl2, although be careful about this if you end up using aoi for other purposes later in the script. I would be interested to hear if the resulting spectra are consistent with your expectations.
Edit for posterity: a github issue has been opened here: https://github.com/pvlib/pvlib-python/issues/1348
Im trying to finetune the existing models in Keras to classify my own dataset. Till now I have tried the following code (taken from Keras docs: https://keras.io/applications/) in which Inception V3 is fine-tuned on a new set of classes.
from keras.applications.inception_v3 import InceptionV3
from keras.preprocessing import image
from keras.models import Model
from keras.layers import Dense, GlobalAveragePooling2D
from keras import backend as K
# create the base pre-trained model
base_model = InceptionV3(weights='imagenet', include_top=False)
# add a global spatial average pooling layer
x = base_model.output
x = GlobalAveragePooling2D()(x)
# let's add a fully-connected layer
x = Dense(1024, activation='relu')(x)
# and a logistic layer -- let's say we have 200 classes
predictions = Dense(200, activation='softmax')(x)
# this is the model we will train
model = Model(inputs=base_model.input, outputs=predictions)
# first: train only the top layers (which were randomly initialized)
# i.e. freeze all convolutional InceptionV3 layers
for layer in base_model.layers:
layer.trainable = False
# compile the model (should be done *after* setting layers to non-trainable)
model.compile(optimizer='rmsprop', loss='categorical_crossentropy')
# train the model on the new data for a few epochs
model.fit_generator(...)
# at this point, the top layers are well trained and we can start fine-tuning
# convolutional layers from inception V3. We will freeze the bottom N layers
# and train the remaining top layers.
# let's visualize layer names and layer indices to see how many layers
# we should freeze:
for i, layer in enumerate(base_model.layers):
print(i, layer.name)
# we chose to train the top 2 inception blocks, i.e. we will freeze
# the first 172 layers and unfreeze the rest:
for layer in model.layers[:172]:
layer.trainable = False
for layer in model.layers[172:]:
layer.trainable = True
# we need to recompile the model for these modifications to take effect
# we use SGD with a low learning rate
from keras.optimizers import SGD
model.compile(optimizer=SGD(lr=0.0001, momentum=0.9), loss='categorical_crossentropy')
# we train our model again (this time fine-tuning the top 2 inception blocks
# alongside the top Dense layers
model.fit_generator(...)
Can anyone plz guide me what changes should I do in the above code so as to fine-tune ResNet50 model present in Keras.
Thanks in advance.
It is difficult to make out a specific question, have you tried anything more than just copying the code without any changes?
That said, there is an abundance of problems in the code: It is a simple copy/paste from keras.io, not functional as it is, and needs some adaption before working at all (regardless of using ResNet50 or InceptionV3):
1): You need to define the input_shape when loading InceptionV3, specifically replace base_model = InceptionV3(weights='imagenet', include_top=False) with base_model = InceptionV3(weights='imagenet', include_top=False, input_shape=(299,299,3))
2): Further, you need to adapt the number of the classes in the last added layer, e.g. if you have only 2 classes to: predictions = Dense(2, activation='softmax')(x)
3): Change the loss-function when compiling your model from categorical_crossentropy to sparse_categorical_crossentropy
4): Most importantly, you need to define the fit_generator before calling model.fit_generator() and add steps_per_epoch. If you have your training images in ./data/train with every category in a different subfolder, this can be done e.g. like this:
from keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator()
train_generator = train_datagen.flow_from_directory(
"./data/train",
target_size=(299, 299),
batch_size=50,
class_mode='binary')
model.fit_generator(train_generator, steps_per_epoch=100)
This of course only does basic training, you will for example need to define save calls to hold on to the trained weights. Only if you get the code working for InceptionV3 with the changes above I suggest to proceed to work on implementing this for ResNet50: As a start you can replace InceptionV3() with ResNet50() (of course only after from keras.applications.resnet50 import ResNet50), and change the input_shape to (224,224,3) and target_size to (224,244).
The above mentioned code-changes should work on Python 3.5.3 / Keras 2.0 / Tensorflow backend.
Beyond the important points mentioned in the above answer for ResNet50 (! if your images are shaped into similar format as in the original Keras code (224,224) - not of rectangular shape) you may substitute:
# add a global spatial average pooling layer
x = base_model.output
x = GlobalAveragePooling2D()(x)
by
x = base_model.output
x = Flatten(x)
EDIT: Please read #Yu-Yang comment bellow
I think I experienced the same issue. It appeared to be a complex problem, which has a decent thread on github(https://github.com/keras-team/keras/issues/9214). The problem is in Batch Normalization of unfreezed blocks of the net. You have two solutions:
Only change top layer(leaving the blocks as they are)
Add a patch from the github thread above.
I am trying to train a RF model in sklearn for classification. The accuracy I get for the test is quite low with a specified set of feature vector. I assume that the feature vector I chose is misleading the model. So I tried RFE, RFECV etc to find a relevant set of feature vector - didn't help to improve the accuracy. I came up with a simple feature selection process as below>
ml_feats = #initial set of feature vector
while True
feats_to_del=[]
prev_score=0
for feat_len in range(2,len(ml_feats)):
classifier = RandomForestClassifier(**init_params)
classifier.fit(X[ml_feats[:feat_len]],Y)
score = classifier.score(Xt[ml_feats[:feat_len]],Yt)
if score<prev_score:
#feature that caused the score to decrease
print ml_feats[feat_len]
feat_to_del.append(ml_feats[feat_len])
prev_score=score
if len(feats_to_del)==0:
break
#delete irrelevant features
ml_feats=list(set(ml_feats)-set(feats_to_del))
print ml_feats #print all relevant features
Does the above code help figure out right set of features?
Thanks
What you are doing is a greedy feature selection. If you want to use RandomForestClassifier to select features, you can do something like:
from sklearn.ensemble import RandomForestClassifier
from sklearn.feature_selection import SelectFromModel
# xtrain : training data
# ytrain : training labels
clf = RandomForestClassifier()
sfm = SelectFromModel(estimator=clf, threshold='mean') # threshold of selection is mean of feature importances by random forest classifier
sfm.fit(xtrain, ytrain)
selected_xtrain = sfm.transform(xtrain)
path_to_jar = path + 'stanford-parser-full-2015-12-09/stanford-parser.jar'
path_to_models_jar = path + 'stanford-parser-full-2015-12-09/stanford-parser-3.6.0-models.jar'
sentence='This is a nice phone 4 me'
print 'Loading module'
start= time.time()
dependency_parser = StanfordDependencyParser(path_to_jar=path_to_jar, path_to_models_jar=path_to_models_jar)
print time.time()-start
start=time.time()
result = dependency_parser.raw_parse((sentence))
print result
print time.time()-start
I've been working on Dependency Parsing using Stanford and NLTK. The probelm I face is the execution time. Here is the output for the above code
Loading module
0.0047550201416
<listiterator object at 0x10bbcbb10>
3.65611600876
It takes approximately 4 seconds per sentence/text. In java, using static variables to load module, its super fast. Any suggestions? At this rate it will take me 100 hours to train provided no error occurs!
I am trying to build a model for the likelihood function of a particular outcome of a Langevin equation (Brownian particle in a harmonic potential):
Here is my model in pymc2 that seems to work:
https://github.com/hstrey/BayesianAnalysis/blob/master/Langevin%20simulation.ipynb
#define the model/function to be fitted.
def model(x):
t = pm.Uniform('t', 0.1, 20, value=2.0)
A = pm.Uniform('A', 0.1, 10, value=1.0)
#pm.deterministic(plot=False)
def S(t=t):
return 1-np.exp(-4*delta_t/t)
#pm.deterministic(plot=False)
def s(t=t):
return np.exp(-2*delta_t/t)
path = np.empty(N, dtype=object)
path[0]=pm.Normal('path_0',mu=0, tau=1/A, value=x[0], observed=True)
for i in range(1,N):
path[i] = pm.Normal('path_%i' % i,
mu=path[i-1]*s,
tau=1/A/S,
value=x[i],
observed=True)
return locals()
mcmc = pm.MCMC( model(x) )
mcmc.sample( 20000, 2000, 10 )
The basic idea is that each point depends on the previous point in the chain (Markov chain). Btw, x is an array of data, N is its length, delta_t is the time step =0.01. Any idea how to implement this in pymc3? I tried:
# define the model/function for diffusion in a harmonic potential
DHP_model = pm.Model()
with DHP_model:
t = pm.Uniform('t', 0.1, 20)
A = pm.Uniform('A', 0.1, 10)
S=1-pm.exp(-4*delta_t/t)
s=pm.exp(-2*delta_t/t)
path = np.empty(N, dtype=object)
path[0]=pm.Normal('path_0',mu=0, tau=1/A, observed=x[0])
for i in range(1,N):
path[i] = pm.Normal('path_%i' % i,
mu=path[i-1]*s,
tau=1/A/S,
observed=x[i])
Unfortunately the model crashes as soon as I try to run it. I tried some pymc3 examples (tutorial) on my machine and this is working.
Thanks in advance. I am really hoping that the new samplers in pymc3 will help me with this model. I am trying to apply Bayesian methods to single-molecule experiments.
Rather than creating many individual normally-distributed 1-D variables in a loop, you can make a custom distribution (by extending Continuous) that knows the formula for computing the log likelihood of your entire path. You can bootstrap this likelihood formula off of the Normal likelihood formula that pymc3 already knows. See the built-in AR1 class for an example.
Since your particle follows the Markov property, your likelihood looks like
import theano.tensor as T
def logp(path):
now = path[1:]
prev = path[:-1]
loglik_first = pm.Normal.dist(mu=0., tau=1./A).logp(path[0])
loglik_rest = T.sum(pm.Normal.dist(mu=prev*ss, tau=1./A/S).logp(now))
loglik_final = loglik_first + loglik_rest
return loglik_final
I'm guessing that you want to draw a value for ss at every time step, in which case you should make sure to specify ss = pm.exp(..., shape=len(x)-1), so that prev*ss in the block above gets interpreted as element-wise multiplication.
Then you can just specify your observations with
path = MyLangevin('path', ..., observed=x)
This should run much faster.
Since I did not see an answer to my question, let me answer it myself. I came up with the following solution:
# now lets model this data using pymc
# define the model/function for diffusion in a harmonic potential
DHP_model = pm.Model()
with DHP_model:
D = pm.Gamma('D',mu=mu_D,sd=sd_D)
A = pm.Gamma('A',mu=mu_A,sd=sd_A)
S=1.0-pm.exp(-2.0*delta_t*D/A)
ss=pm.exp(-delta_t*D/A)
path=pm.Normal('path_0',mu=0.0, tau=1/A, observed=x[0])
for i in range(1,N):
path = pm.Normal('path_%i' % i,
mu=path*ss,
tau=1.0/A/S,
observed=x[i])
start = pm.find_MAP()
print(start)
trace = pm.sample(100000, start=start)
unfortunately, this code takes at N=50 anywhere between 6hours to 2 days to compile. I am running on a pretty fast PC (24Gb RAM) running Ubuntu. I tried to using the GPU but that runs slightly slower. I suspect memory problems since it uses 99.8% of the memory when running. I tried the same calculation with Stan and it only takes 2min to run.