mirror of
https://github.com/titanscouting/tra-analysis.git
synced 2024-11-10 06:54:44 +00:00
depreciated 2019 superscripts and company
This commit is contained in:
art
parent
2bdb15a2b3
commit
9f0d366408
@ -1,61 +1,61 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Created on Wed Mar 20 12:21:31 2019
|
||||
|
||||
@author: creek
|
||||
"""
|
||||
import firebase_admin
|
||||
from firebase_admin import credentials
|
||||
from firebase_admin import firestore
|
||||
import pprint
|
||||
from pylatex import Document, Section, Subsection, Command
|
||||
from pylatex.utils import italic, NoEscape
|
||||
import requests
|
||||
|
||||
def generate_team_report(team):
|
||||
doc = Document('basic')
|
||||
matches = team.reference.collection(u'matches').get()
|
||||
matchnums = []
|
||||
for match in matches:
|
||||
matchnums.append(match.id)
|
||||
with doc.create(Section('Qualification matches scouted')):
|
||||
for matchnum in matchnums:
|
||||
doc.append(str(matchnum))
|
||||
with doc.create(Section('Details')):
|
||||
hab = "Hab 1"
|
||||
balls = 42
|
||||
hatches = 0
|
||||
count = 0
|
||||
for match in matches:
|
||||
for analysis in match:
|
||||
if analysis.key().startswith('Quant'):
|
||||
balls = balls + analysis['cargoBalls']
|
||||
hatches = hatches + analysis['hatchPanels']
|
||||
count = count + 1
|
||||
if analysis.key().startswith('Qual'):
|
||||
strategy = analysis['StrategyType']
|
||||
strongObject = analysis['TeleopStrongObject']
|
||||
if count > 0:
|
||||
doc.append("Average balls: " + str(float(balls)/count))
|
||||
doc.append("Average hatches: " + str(float(hatches)/count))
|
||||
doc.append("Strategy Type: " + str(strategy))
|
||||
doc.append("Strongest object in teleop: " + str(strongObject))
|
||||
|
||||
|
||||
doc.preamble.append(Command('title', team.id))
|
||||
doc.preamble.append(Command('author', 'Generated by Team 2022'))
|
||||
doc.preamble.append(Command('date', NoEscape(r'\today')))
|
||||
doc.append(NoEscape(r'\maketitle'))
|
||||
|
||||
doc.generate_pdf(filepath= str(team.id), clean_tex=False)
|
||||
|
||||
credential = credentials.Certificate('keys/firebasekey.json')
|
||||
|
||||
firebase_admin.initialize_app(credential)
|
||||
|
||||
db = firestore.Client()
|
||||
teams_ref = db.collection(u'data').document(u'team-2022').collection(u'Central 2019')
|
||||
teams = teams_ref.get()
|
||||
|
||||
for team in teams:
|
||||
generate_team_report(team)
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Created on Wed Mar 20 12:21:31 2019
|
||||
|
||||
@author: creek
|
||||
"""
|
||||
import firebase_admin
|
||||
from firebase_admin import credentials
|
||||
from firebase_admin import firestore
|
||||
import pprint
|
||||
from pylatex import Document, Section, Subsection, Command
|
||||
from pylatex.utils import italic, NoEscape
|
||||
import requests
|
||||
|
||||
def generate_team_report(team):
|
||||
doc = Document('basic')
|
||||
matches = team.reference.collection(u'matches').get()
|
||||
matchnums = []
|
||||
for match in matches:
|
||||
matchnums.append(match.id)
|
||||
with doc.create(Section('Qualification matches scouted')):
|
||||
for matchnum in matchnums:
|
||||
doc.append(str(matchnum))
|
||||
with doc.create(Section('Details')):
|
||||
hab = "Hab 1"
|
||||
balls = 42
|
||||
hatches = 0
|
||||
count = 0
|
||||
for match in matches:
|
||||
for analysis in match:
|
||||
if analysis.key().startswith('Quant'):
|
||||
balls = balls + analysis['cargoBalls']
|
||||
hatches = hatches + analysis['hatchPanels']
|
||||
count = count + 1
|
||||
if analysis.key().startswith('Qual'):
|
||||
strategy = analysis['StrategyType']
|
||||
strongObject = analysis['TeleopStrongObject']
|
||||
if count > 0:
|
||||
doc.append("Average balls: " + str(float(balls)/count))
|
||||
doc.append("Average hatches: " + str(float(hatches)/count))
|
||||
doc.append("Strategy Type: " + str(strategy))
|
||||
doc.append("Strongest object in teleop: " + str(strongObject))
|
||||
|
||||
|
||||
doc.preamble.append(Command('title', team.id))
|
||||
doc.preamble.append(Command('author', 'Generated by Team 2022'))
|
||||
doc.preamble.append(Command('date', NoEscape(r'\today')))
|
||||
doc.append(NoEscape(r'\maketitle'))
|
||||
|
||||
doc.generate_pdf(filepath= str(team.id), clean_tex=False)
|
||||
|
||||
credential = credentials.Certificate('keys/firebasekey.json')
|
||||
|
||||
firebase_admin.initialize_app(credential)
|
||||
|
||||
db = firestore.Client()
|
||||
teams_ref = db.collection(u'data').document(u'team-2022').collection(u'Central 2019')
|
||||
teams = teams_ref.get()
|
||||
|
||||
for team in teams:
|
||||
generate_team_report(team)
|
||||
@ -1,206 +1,206 @@
|
||||
#Titan Robotics Team 2022: ML Module
|
||||
#Written by Arthur Lu & Jacob Levine
|
||||
#Notes:
|
||||
# this should be imported as a python module using 'import titanlearn'
|
||||
# this should be included in the local directory or environment variable
|
||||
# this module has not been optimized for multhreaded computing
|
||||
# this module learns from its mistakes far faster than 2022's captains
|
||||
#setup:
|
||||
|
||||
__version__ = "1.0.0.001"
|
||||
|
||||
#changelog should be viewed using print(analysis.__changelog__)
|
||||
__changelog__ = """changelog:
|
||||
1.0.0.xxx:
|
||||
-added generation of ANNS, basic SGD training"""
|
||||
__author__ = (
|
||||
"Arthur Lu <arthurlu@ttic.edu>, "
|
||||
"Jacob Levine <jlevine@ttic.edu>,"
|
||||
)
|
||||
__all__ = [
|
||||
'linear_nn',
|
||||
'train_sgd_minibatch',
|
||||
'train_sgd_simple'
|
||||
]
|
||||
#imports
|
||||
import torch
|
||||
import warnings
|
||||
from collections import OrderedDict
|
||||
from sklearn import metrics, datasets
|
||||
import numpy as np
|
||||
import matplotlib.pyplot as plt
|
||||
import math
|
||||
import time
|
||||
|
||||
#enable CUDA if possible
|
||||
device = torch.device("cpu")
|
||||
|
||||
#linear_nn: creates a fully connected network given params
|
||||
def linear_nn(in_dim, hidden_dim, out_dim, num_hidden, act_fn="tanh", end="none"):
|
||||
if act_fn.lower()=="tanh":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "tanh"+str(i+1):torch.nn.Tanh()})
|
||||
|
||||
elif act_fn.lower()=="sigmoid":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "sig"+str(i+1):torch.nn.Sigmoid()})
|
||||
|
||||
elif act_fn.lower()=="relu":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "relu"+str(i+1):torch.nn.ReLU()})
|
||||
|
||||
elif act_fn.lower()=="leaky relu":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "lre"+str(i+1):torch.nn.LeakyReLU()})
|
||||
else:
|
||||
warnings.warn("Did not specify a valid inner activation function. Returning nothing.")
|
||||
return None
|
||||
|
||||
if end.lower()=="softmax":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim), "softmax": torch.nn.Softmax()})
|
||||
elif end.lower()=="none":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim)})
|
||||
elif end.lower()=="sigmoid":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim), "sigmoid": torch.nn.Sigmoid()})
|
||||
else:
|
||||
warnings.warn("Did not specify a valid final activation function. Returning nothing.")
|
||||
return None
|
||||
|
||||
return torch.nn.Sequential(k)
|
||||
|
||||
#train_sgd_simple: trains network using SGD
|
||||
def train_sgd_simple(net, evalType, data, ground, dev=None, devg=None, iters=1000, learnrate=1e-4, testevery=1, graphsaveloc=None, modelsaveloc=None, loss="mse"):
|
||||
model=net.to(device)
|
||||
data=data.to(device)
|
||||
ground=ground.to(device)
|
||||
if dev != None:
|
||||
dev=dev.to(device)
|
||||
losses=[]
|
||||
dev_losses=[]
|
||||
if loss.lower()=="mse":
|
||||
loss_fn = torch.nn.MSELoss()
|
||||
elif loss.lower()=="cross entropy":
|
||||
loss_fn = torch.nn.CrossEntropyLoss()
|
||||
elif loss.lower()=="nll":
|
||||
loss_fn = torch.nn.NLLLoss()
|
||||
elif loss.lower()=="poisson nll":
|
||||
loss_fn = torch.nn.PoissonNLLLoss()
|
||||
else:
|
||||
warnings.warn("Did not specify a valid loss function. Returning nothing.")
|
||||
return None
|
||||
optimizer=torch.optim.SGD(model.parameters(), lr=learnrate)
|
||||
for i in range(iters):
|
||||
if i%testevery==0:
|
||||
with torch.no_grad():
|
||||
output = model(data)
|
||||
if evalType == "ap":
|
||||
ap = metrics.average_precision_score(ground.cpu().numpy(), output.cpu().numpy())
|
||||
if evalType == "regression":
|
||||
ap = metrics.explained_variance_score(ground.cpu().numpy(), output.cpu().numpy())
|
||||
losses.append(ap)
|
||||
print(str(i)+": "+str(ap))
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="train AP")
|
||||
if dev != None:
|
||||
output = model(dev)
|
||||
print(evalType)
|
||||
if evalType == "ap":
|
||||
|
||||
ap = metrics.average_precision_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ap)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev AP")
|
||||
elif evalType == "regression":
|
||||
ev = metrics.explained_variance_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ev)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev EV")
|
||||
|
||||
|
||||
if graphsaveloc != None:
|
||||
plt.savefig(graphsaveloc+".pdf")
|
||||
with torch.enable_grad():
|
||||
optimizer.zero_grad()
|
||||
output = model(data)
|
||||
loss = loss_fn(output, ground)
|
||||
print(loss.item())
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
if modelsaveloc != None:
|
||||
torch.save(model, modelsaveloc)
|
||||
plt.show()
|
||||
return model
|
||||
|
||||
#train_sgd_minibatch: same as above, but with minibatches
|
||||
def train_sgd_minibatch(net, data, ground, dev=None, devg=None, epoch=100, batchsize=20, learnrate=1e-4, testevery=20, graphsaveloc=None, modelsaveloc=None, loss="mse"):
|
||||
model=net.to(device)
|
||||
data=data.to(device)
|
||||
ground=ground.to(device)
|
||||
if dev != None:
|
||||
dev=dev.to(device)
|
||||
losses=[]
|
||||
dev_losses=[]
|
||||
if loss.lower()=="mse":
|
||||
loss_fn = torch.nn.MSELoss()
|
||||
elif loss.lower()=="cross entropy":
|
||||
loss_fn = torch.nn.CrossEntropyLoss()
|
||||
elif loss.lower()=="nll":
|
||||
loss_fn = torch.nn.NLLLoss()
|
||||
elif loss.lower()=="poisson nll":
|
||||
loss_fn = torch.nn.PoissonNLLLoss()
|
||||
else:
|
||||
warnings.warn("Did not specify a valid loss function. Returning nothing.")
|
||||
return None
|
||||
optimizer=torch.optim.LBFGS(model.parameters(), lr=learnrate)
|
||||
itercount=0
|
||||
for i in range(epoch):
|
||||
print("EPOCH "+str(i)+" OF "+str(epoch-1))
|
||||
batches=math.ceil(data.size()[0].item()/batchsize)
|
||||
for j in range(batches):
|
||||
batchdata=[]
|
||||
batchground=[]
|
||||
for k in range(j*batchsize, min((j+1)*batchsize, data.size()[0].item()),1):
|
||||
batchdata.append(data[k])
|
||||
batchground.append(ground[k])
|
||||
batchdata=torch.stack(batchdata)
|
||||
batchground=torch.stack(batchground)
|
||||
if itercount%testevery==0:
|
||||
with torch.no_grad():
|
||||
output = model(data)
|
||||
ap = metrics.average_precision_score(ground.numpy(), output.numpy())
|
||||
losses.append(ap)
|
||||
print(str(i)+": "+str(ap))
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses))
|
||||
if dev != None:
|
||||
output = model(dev)
|
||||
ap = metrics.average_precision_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ap)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev AP")
|
||||
if graphsaveloc != None:
|
||||
plt.savefig(graphsaveloc+".pdf")
|
||||
with torch.enable_grad():
|
||||
optimizer.zero_grad()
|
||||
output = model(batchdata)
|
||||
loss = loss_fn(output, ground)
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
itercount +=1
|
||||
if modelsaveloc != None:
|
||||
torch.save(model, modelsaveloc)
|
||||
plt.show()
|
||||
return model
|
||||
|
||||
def retyuoipufdyu():
|
||||
|
||||
data = torch.tensor(datasets.fetch_california_housing()['data']).to(torch.float)
|
||||
ground = datasets.fetch_california_housing()['target']
|
||||
ground = torch.tensor(ground).to(torch.float)
|
||||
model = linear_nn(8, 100, 1, 20, act_fn = "relu")
|
||||
print(model)
|
||||
return train_sgd_simple(model,"regression", data, ground, learnrate=1e-4, iters=1000)
|
||||
|
||||
start = time.time()
|
||||
retyuoipufdyu()
|
||||
end = time.time()
|
||||
print(end-start)
|
||||
#Titan Robotics Team 2022: ML Module
|
||||
#Written by Arthur Lu & Jacob Levine
|
||||
#Notes:
|
||||
# this should be imported as a python module using 'import titanlearn'
|
||||
# this should be included in the local directory or environment variable
|
||||
# this module has not been optimized for multhreaded computing
|
||||
# this module learns from its mistakes far faster than 2022's captains
|
||||
#setup:
|
||||
|
||||
__version__ = "1.0.0.001"
|
||||
|
||||
#changelog should be viewed using print(analysis.__changelog__)
|
||||
__changelog__ = """changelog:
|
||||
1.0.0.xxx:
|
||||
-added generation of ANNS, basic SGD training"""
|
||||
__author__ = (
|
||||
"Arthur Lu <arthurlu@ttic.edu>, "
|
||||
"Jacob Levine <jlevine@ttic.edu>,"
|
||||
)
|
||||
__all__ = [
|
||||
'linear_nn',
|
||||
'train_sgd_minibatch',
|
||||
'train_sgd_simple'
|
||||
]
|
||||
#imports
|
||||
import torch
|
||||
import warnings
|
||||
from collections import OrderedDict
|
||||
from sklearn import metrics, datasets
|
||||
import numpy as np
|
||||
import matplotlib.pyplot as plt
|
||||
import math
|
||||
import time
|
||||
|
||||
#enable CUDA if possible
|
||||
device = torch.device("cpu")
|
||||
|
||||
#linear_nn: creates a fully connected network given params
|
||||
def linear_nn(in_dim, hidden_dim, out_dim, num_hidden, act_fn="tanh", end="none"):
|
||||
if act_fn.lower()=="tanh":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "tanh"+str(i+1):torch.nn.Tanh()})
|
||||
|
||||
elif act_fn.lower()=="sigmoid":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "sig"+str(i+1):torch.nn.Sigmoid()})
|
||||
|
||||
elif act_fn.lower()=="relu":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "relu"+str(i+1):torch.nn.ReLU()})
|
||||
|
||||
elif act_fn.lower()=="leaky relu":
|
||||
k=OrderedDict([("in", torch.nn.Linear(in_dim,hidden_dim))])
|
||||
for i in range(num_hidden):
|
||||
k.update({"lin"+str(i+1): torch.nn.Linear(hidden_dim,hidden_dim), "lre"+str(i+1):torch.nn.LeakyReLU()})
|
||||
else:
|
||||
warnings.warn("Did not specify a valid inner activation function. Returning nothing.")
|
||||
return None
|
||||
|
||||
if end.lower()=="softmax":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim), "softmax": torch.nn.Softmax()})
|
||||
elif end.lower()=="none":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim)})
|
||||
elif end.lower()=="sigmoid":
|
||||
k.update({"out": torch.nn.Linear(hidden_dim,out_dim), "sigmoid": torch.nn.Sigmoid()})
|
||||
else:
|
||||
warnings.warn("Did not specify a valid final activation function. Returning nothing.")
|
||||
return None
|
||||
|
||||
return torch.nn.Sequential(k)
|
||||
|
||||
#train_sgd_simple: trains network using SGD
|
||||
def train_sgd_simple(net, evalType, data, ground, dev=None, devg=None, iters=1000, learnrate=1e-4, testevery=1, graphsaveloc=None, modelsaveloc=None, loss="mse"):
|
||||
model=net.to(device)
|
||||
data=data.to(device)
|
||||
ground=ground.to(device)
|
||||
if dev != None:
|
||||
dev=dev.to(device)
|
||||
losses=[]
|
||||
dev_losses=[]
|
||||
if loss.lower()=="mse":
|
||||
loss_fn = torch.nn.MSELoss()
|
||||
elif loss.lower()=="cross entropy":
|
||||
loss_fn = torch.nn.CrossEntropyLoss()
|
||||
elif loss.lower()=="nll":
|
||||
loss_fn = torch.nn.NLLLoss()
|
||||
elif loss.lower()=="poisson nll":
|
||||
loss_fn = torch.nn.PoissonNLLLoss()
|
||||
else:
|
||||
warnings.warn("Did not specify a valid loss function. Returning nothing.")
|
||||
return None
|
||||
optimizer=torch.optim.SGD(model.parameters(), lr=learnrate)
|
||||
for i in range(iters):
|
||||
if i%testevery==0:
|
||||
with torch.no_grad():
|
||||
output = model(data)
|
||||
if evalType == "ap":
|
||||
ap = metrics.average_precision_score(ground.cpu().numpy(), output.cpu().numpy())
|
||||
if evalType == "regression":
|
||||
ap = metrics.explained_variance_score(ground.cpu().numpy(), output.cpu().numpy())
|
||||
losses.append(ap)
|
||||
print(str(i)+": "+str(ap))
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="train AP")
|
||||
if dev != None:
|
||||
output = model(dev)
|
||||
print(evalType)
|
||||
if evalType == "ap":
|
||||
|
||||
ap = metrics.average_precision_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ap)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev AP")
|
||||
elif evalType == "regression":
|
||||
ev = metrics.explained_variance_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ev)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev EV")
|
||||
|
||||
|
||||
if graphsaveloc != None:
|
||||
plt.savefig(graphsaveloc+".pdf")
|
||||
with torch.enable_grad():
|
||||
optimizer.zero_grad()
|
||||
output = model(data)
|
||||
loss = loss_fn(output, ground)
|
||||
print(loss.item())
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
if modelsaveloc != None:
|
||||
torch.save(model, modelsaveloc)
|
||||
plt.show()
|
||||
return model
|
||||
|
||||
#train_sgd_minibatch: same as above, but with minibatches
|
||||
def train_sgd_minibatch(net, data, ground, dev=None, devg=None, epoch=100, batchsize=20, learnrate=1e-4, testevery=20, graphsaveloc=None, modelsaveloc=None, loss="mse"):
|
||||
model=net.to(device)
|
||||
data=data.to(device)
|
||||
ground=ground.to(device)
|
||||
if dev != None:
|
||||
dev=dev.to(device)
|
||||
losses=[]
|
||||
dev_losses=[]
|
||||
if loss.lower()=="mse":
|
||||
loss_fn = torch.nn.MSELoss()
|
||||
elif loss.lower()=="cross entropy":
|
||||
loss_fn = torch.nn.CrossEntropyLoss()
|
||||
elif loss.lower()=="nll":
|
||||
loss_fn = torch.nn.NLLLoss()
|
||||
elif loss.lower()=="poisson nll":
|
||||
loss_fn = torch.nn.PoissonNLLLoss()
|
||||
else:
|
||||
warnings.warn("Did not specify a valid loss function. Returning nothing.")
|
||||
return None
|
||||
optimizer=torch.optim.LBFGS(model.parameters(), lr=learnrate)
|
||||
itercount=0
|
||||
for i in range(epoch):
|
||||
print("EPOCH "+str(i)+" OF "+str(epoch-1))
|
||||
batches=math.ceil(data.size()[0].item()/batchsize)
|
||||
for j in range(batches):
|
||||
batchdata=[]
|
||||
batchground=[]
|
||||
for k in range(j*batchsize, min((j+1)*batchsize, data.size()[0].item()),1):
|
||||
batchdata.append(data[k])
|
||||
batchground.append(ground[k])
|
||||
batchdata=torch.stack(batchdata)
|
||||
batchground=torch.stack(batchground)
|
||||
if itercount%testevery==0:
|
||||
with torch.no_grad():
|
||||
output = model(data)
|
||||
ap = metrics.average_precision_score(ground.numpy(), output.numpy())
|
||||
losses.append(ap)
|
||||
print(str(i)+": "+str(ap))
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses))
|
||||
if dev != None:
|
||||
output = model(dev)
|
||||
ap = metrics.average_precision_score(devg.numpy(), output.numpy())
|
||||
dev_losses.append(ap)
|
||||
plt.plot(np.array(range(0,i+1,testevery)),np.array(losses), label="dev AP")
|
||||
if graphsaveloc != None:
|
||||
plt.savefig(graphsaveloc+".pdf")
|
||||
with torch.enable_grad():
|
||||
optimizer.zero_grad()
|
||||
output = model(batchdata)
|
||||
loss = loss_fn(output, ground)
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
itercount +=1
|
||||
if modelsaveloc != None:
|
||||
torch.save(model, modelsaveloc)
|
||||
plt.show()
|
||||
return model
|
||||
|
||||
def retyuoipufdyu():
|
||||
|
||||
data = torch.tensor(datasets.fetch_california_housing()['data']).to(torch.float)
|
||||
ground = datasets.fetch_california_housing()['target']
|
||||
ground = torch.tensor(ground).to(torch.float)
|
||||
model = linear_nn(8, 100, 1, 20, act_fn = "relu")
|
||||
print(model)
|
||||
return train_sgd_simple(model,"regression", data, ground, learnrate=1e-4, iters=1000)
|
||||
|
||||
start = time.time()
|
||||
retyuoipufdyu()
|
||||
end = time.time()
|
||||
print(end-start)
|
||||
@ -1,130 +1,130 @@
|
||||
#Titan Robotics Team 2022: Visualization Module
|
||||
#Written by Arthur Lu & Jacob Levine
|
||||
#Notes:
|
||||
# this should be imported as a python module using 'import visualization'
|
||||
# this should be included in the local directory or environment variable
|
||||
# this module has not been optimized for multhreaded computing
|
||||
#Number of easter eggs: Jake is Jewish and does not observe easter.
|
||||
#setup:
|
||||
|
||||
__version__ = "1.0.0.001"
|
||||
|
||||
#changelog should be viewed using print(analysis.__changelog__)
|
||||
__changelog__ = """changelog:
|
||||
1.0.0.xxx:
|
||||
-added basic plotting, clustering, and regression comparisons"""
|
||||
__author__ = (
|
||||
"Arthur Lu <arthurlu@ttic.edu>, "
|
||||
"Jacob Levine <jlevine@ttic.edu>,"
|
||||
)
|
||||
__all__ = [
|
||||
'affinity_prop',
|
||||
'bar_graph',
|
||||
'dbscan',
|
||||
'kmeans',
|
||||
'line_plot',
|
||||
'pca_comp',
|
||||
'regression_comp',
|
||||
'scatter_plot',
|
||||
'spectral',
|
||||
'vis_2d'
|
||||
]
|
||||
#imports
|
||||
import matplotlib.pyplot as plt
|
||||
import numpy as np
|
||||
from sklearn.decomposition import PCA, KernelPCA, IncrementalPCA
|
||||
from sklearn.preprocessing import StandardScaler
|
||||
from sklearn.cluster import AffinityPropagation, DBSCAN, KMeans, SpectralClustering
|
||||
|
||||
#bar of x,y
|
||||
def bar_graph(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.bar(x,y)
|
||||
plt.show()
|
||||
|
||||
#scatter of x,y
|
||||
def scatter_plot(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.scatter(x,y)
|
||||
plt.show()
|
||||
|
||||
#line of x,y
|
||||
def line_plot(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.scatter(x,y)
|
||||
plt.show()
|
||||
|
||||
#plot data + regression fit
|
||||
def regression_comp(x,y,reg):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
regx=np.arange(x.min(),x.max(),(x.max()-x.min())/1000)
|
||||
regy=[]
|
||||
for i in regx:
|
||||
regy.append(eval(reg[0].replace("z",str(i))))
|
||||
regy=np.asarray(regy)
|
||||
plt.scatter(x,y)
|
||||
plt.plot(regx,regy,color="orange",linewidth=3)
|
||||
plt.text(.85*max([x.max(),regx.max()]),.95*max([y.max(),regy.max()]),
|
||||
u"R\u00b2="+str(round(reg[2],5)),
|
||||
horizontalalignment='center', verticalalignment='center')
|
||||
plt.text(.85*max([x.max(),regx.max()]),.85*max([y.max(),regy.max()]),
|
||||
"MSE="+str(round(reg[1],5)),
|
||||
horizontalalignment='center', verticalalignment='center')
|
||||
plt.show()
|
||||
|
||||
#PCA to compress down to 2d
|
||||
def pca_comp(big_multidim):
|
||||
pca=PCA(n_components=2)
|
||||
td_norm=StandardScaler().fit_transform(big_multidim)
|
||||
td_pca=pca.fit_transform(td_norm)
|
||||
return td_pca
|
||||
|
||||
#one-stop visualization of multidim datasets
|
||||
def vis_2d(big_multidim):
|
||||
td_pca=pca_comp(big_multidim)
|
||||
plt.scatter(td_pca[:,0], td_pca[:,1])
|
||||
|
||||
def cluster_vis(data, cluster_assign):
|
||||
pca=PCA(n_components=2)
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
td_pca=pca.fit_transform(td_norm)
|
||||
colors = np.array(list(islice(cycle(['#377eb8', '#ff7f00', '#4daf4a',
|
||||
'#f781bf', '#a65628', '#984ea3',
|
||||
'#999999', '#e41a1c', '#dede00']),
|
||||
int(max(clu) + 1))))
|
||||
colors = np.append(colors, ["#000000"])
|
||||
plt.figure(figsize=(8, 8))
|
||||
plt.scatter(td_norm[:, 0], td_norm[:, 1], s=10, color=colors[cluster_assign])
|
||||
plt.show()
|
||||
|
||||
#affinity prop- slow, but ok if you don't have any idea how many you want
|
||||
def affinity_prop(data, damping=.77, preference=-70):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = AffinityPropagation(damping=damping,preference=preference).fit(td)
|
||||
y=db.predict(td_norm)
|
||||
return y
|
||||
|
||||
#DBSCAN- slightly faster but can label your dataset as all outliers
|
||||
def dbscan(data, eps=.3):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = DBSCAN(eps=eps).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
|
||||
#K-means clustering- the classic
|
||||
def kmeans(data, num_clusters):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = KMeans(n_clusters=num_clusters).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
|
||||
#Spectral Clustering- Seems to work really well
|
||||
def spectral(data, num_clusters):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = SpectralClustering(n_clusters=num_clusters).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
#Titan Robotics Team 2022: Visualization Module
|
||||
#Written by Arthur Lu & Jacob Levine
|
||||
#Notes:
|
||||
# this should be imported as a python module using 'import visualization'
|
||||
# this should be included in the local directory or environment variable
|
||||
# this module has not been optimized for multhreaded computing
|
||||
#Number of easter eggs: Jake is Jewish and does not observe easter.
|
||||
#setup:
|
||||
|
||||
__version__ = "1.0.0.001"
|
||||
|
||||
#changelog should be viewed using print(analysis.__changelog__)
|
||||
__changelog__ = """changelog:
|
||||
1.0.0.xxx:
|
||||
-added basic plotting, clustering, and regression comparisons"""
|
||||
__author__ = (
|
||||
"Arthur Lu <arthurlu@ttic.edu>, "
|
||||
"Jacob Levine <jlevine@ttic.edu>,"
|
||||
)
|
||||
__all__ = [
|
||||
'affinity_prop',
|
||||
'bar_graph',
|
||||
'dbscan',
|
||||
'kmeans',
|
||||
'line_plot',
|
||||
'pca_comp',
|
||||
'regression_comp',
|
||||
'scatter_plot',
|
||||
'spectral',
|
||||
'vis_2d'
|
||||
]
|
||||
#imports
|
||||
import matplotlib.pyplot as plt
|
||||
import numpy as np
|
||||
from sklearn.decomposition import PCA, KernelPCA, IncrementalPCA
|
||||
from sklearn.preprocessing import StandardScaler
|
||||
from sklearn.cluster import AffinityPropagation, DBSCAN, KMeans, SpectralClustering
|
||||
|
||||
#bar of x,y
|
||||
def bar_graph(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.bar(x,y)
|
||||
plt.show()
|
||||
|
||||
#scatter of x,y
|
||||
def scatter_plot(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.scatter(x,y)
|
||||
plt.show()
|
||||
|
||||
#line of x,y
|
||||
def line_plot(x,y):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
plt.scatter(x,y)
|
||||
plt.show()
|
||||
|
||||
#plot data + regression fit
|
||||
def regression_comp(x,y,reg):
|
||||
x=np.asarray(x)
|
||||
y=np.asarray(y)
|
||||
regx=np.arange(x.min(),x.max(),(x.max()-x.min())/1000)
|
||||
regy=[]
|
||||
for i in regx:
|
||||
regy.append(eval(reg[0].replace("z",str(i))))
|
||||
regy=np.asarray(regy)
|
||||
plt.scatter(x,y)
|
||||
plt.plot(regx,regy,color="orange",linewidth=3)
|
||||
plt.text(.85*max([x.max(),regx.max()]),.95*max([y.max(),regy.max()]),
|
||||
u"R\u00b2="+str(round(reg[2],5)),
|
||||
horizontalalignment='center', verticalalignment='center')
|
||||
plt.text(.85*max([x.max(),regx.max()]),.85*max([y.max(),regy.max()]),
|
||||
"MSE="+str(round(reg[1],5)),
|
||||
horizontalalignment='center', verticalalignment='center')
|
||||
plt.show()
|
||||
|
||||
#PCA to compress down to 2d
|
||||
def pca_comp(big_multidim):
|
||||
pca=PCA(n_components=2)
|
||||
td_norm=StandardScaler().fit_transform(big_multidim)
|
||||
td_pca=pca.fit_transform(td_norm)
|
||||
return td_pca
|
||||
|
||||
#one-stop visualization of multidim datasets
|
||||
def vis_2d(big_multidim):
|
||||
td_pca=pca_comp(big_multidim)
|
||||
plt.scatter(td_pca[:,0], td_pca[:,1])
|
||||
|
||||
def cluster_vis(data, cluster_assign):
|
||||
pca=PCA(n_components=2)
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
td_pca=pca.fit_transform(td_norm)
|
||||
colors = np.array(list(islice(cycle(['#377eb8', '#ff7f00', '#4daf4a',
|
||||
'#f781bf', '#a65628', '#984ea3',
|
||||
'#999999', '#e41a1c', '#dede00']),
|
||||
int(max(clu) + 1))))
|
||||
colors = np.append(colors, ["#000000"])
|
||||
plt.figure(figsize=(8, 8))
|
||||
plt.scatter(td_norm[:, 0], td_norm[:, 1], s=10, color=colors[cluster_assign])
|
||||
plt.show()
|
||||
|
||||
#affinity prop- slow, but ok if you don't have any idea how many you want
|
||||
def affinity_prop(data, damping=.77, preference=-70):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = AffinityPropagation(damping=damping,preference=preference).fit(td)
|
||||
y=db.predict(td_norm)
|
||||
return y
|
||||
|
||||
#DBSCAN- slightly faster but can label your dataset as all outliers
|
||||
def dbscan(data, eps=.3):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = DBSCAN(eps=eps).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
|
||||
#K-means clustering- the classic
|
||||
def kmeans(data, num_clusters):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = KMeans(n_clusters=num_clusters).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
|
||||
#Spectral Clustering- Seems to work really well
|
||||
def spectral(data, num_clusters):
|
||||
td_norm=StandardScaler().fit_transform(data)
|
||||
db = SpectralClustering(n_clusters=num_clusters).fit(td)
|
||||
y=db.labels_.astype(np.int)
|
||||
return y
|
||||
Loading…
Reference in New Issue
Block a user