#%% Create DataLoader for First PSM
file_dir = '../PSM1_data' # define the file directory for dataset
model_type = "S"
train_list = [1,3,5,7]
val_list = [1]
test_list = [1]
config_dict={'file_dir':file_dir,
'include_torque': False,
'spatial_forces': False,
'custom_state': None,
'batch_size': 32,
'crop_list':None,
'trans_function': None}
loader_dict,loader_sizes = dat.init_dataset(train_list,val_list,test_list,model_type,config_dict)
# from the loader_dict extract the mean and std of the train set
mean1 = loader_dict["train"].dataset.mean
std1 = loader_dict["train"].dataset.stdev
#%% Create DataLoader for the Second PSM
file_dir = '../experiment_data' # define the file directory for dataset
model_type = "S"
train_list = [1,3,5,7,
8,10,12,14,
15,17,19,21]
val_list = [1]
test_list = [1]
config_dict={'file_dir':file_dir,
train_list = [1, 3, 5, 7, 8, 10, 12, 14, 15, 17, 19, 21]
val_list = [2, 6, 9, 13, 16, 20]
test_list = [4, 11, 18, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33]
config_dict = {
'file_dir': file_dir,
'include_torque': False,
'spatial_forces': force_align,
'custom_state': None,
'batch_size': 32,
'crop_list': crop_list,
'trans_function': trans_function
}
dataloaders, dataset_sizes = dat.init_dataset(train_list, val_list,
val_list, model_type,
config_dict)
# set the logarithmic learning rate
learning_rates = [1e-1, 1e-2, 1e-3, 1e-4, 1e-5, 1e-6]
best_loss = np.inf
best_lr = 0
for lr in learning_rates:
print('lr:{}'.format(lr))
# define model
if model_type == "VS":
model = mdl.StateVisionModel(30,
54,
3,
feature_extract=feat_extract)
#test_list = [4,11,18,
#22,23,24,25,26,27,28,29,32,33]
test_list = [4, 8]
config_dict = {
'file_dir': file_dir,
'include_torque': False,
'spatial_forces': force_align,
'custom_state': None,
'batch_size': 32,
'crop_list': crop_list,
'trans_function': trans_function
}
dataloaders, dataset_sizes = dat.init_dataset(train_list,
val_list,
test_list,
model_type,
config_dict,
augment=False)
np.savetxt('PSM2_mean_smalldata.csv', dataloaders['train'].dataset.mean)
np.savetxt('PSM2_std_smalldata.csv', dataloaders['train'].dataset.stdev)
'''
## if we ablate uncomment these lines -----------------------------
qty = ['t','fx','fy','fz','tx','ty','tz',
'px','py','pz','qx','qy','qz','qw','vx','vy','vz','wx','wy','wz',
'q1','q2','q3','q4','q5','q6','q7',
'vq1','vq2','vq3','vq4','vq5','vq6','vq7',
'tq1','tq2','tq3','tq4','tq5','tq6','tq7',
'q1d','q2d','q3d','q4d','q5d','q6d','q7d',
'tq1d','tq2d','tq3d','tq4d','tq5d','tq6d','tq7d',
'psm_fx','psm_fy','psm_fz','psm_tx','psm_ty','psm_tz',
'J1','J2','J3','J4','J5','J6','J1','J2','J3','J4','J5','J6',
def run_ablations(model, num_ablations):
'''set up some persistent tracking variables'''
remove_features = [] # list of features we are removing
metrics_list = [] # list storing dictionary of performance metrics
# feature indexes
full_state_index = np.arange(7, 61)
input_state = 54
# create loss function
criterion = nn.MSELoss(reduction='sum')
# define optimization method
optimizer = opt.Adam(model.parameters(), lr=0.01)
param_count = []
param_count.append(count_params(model))
current_feature_list = np.array(qty)
# create the dataloader
dataloaders, dataset_sizes = dat.init_dataset(train_list, val_list,
val_list, model_type,
config_dict)
print('evaluating full model predictions...')
predictions = mdl.evaluate_model(model,
dataloaders['test'],
model_type=model_type,
no_pbar=True)
# compute the loss statistics
print('computing full model performance metrics...')
metrics = model_eval.compute_loss_metrics(
predictions, dataloaders['test'].dataset.label_array[:, 1:4])
metrics_list.append(metrics)
print('Performance Summary of Full Model:')
print(metrics)
print('Running ablation study on model type:' + model_type)
for iteration in range(num_ablations):
print('-' * 10)
print('Begin ablation run: {}/{}'.format(iteration + 1, num_ablations))
print('-' * 10)
# compute the backprop values:
gbp_data = model_eval.compute_GBP(model,
dataloaders['test'],
num_state_inputs=input_state,
model_type=model_type,
no_pbar=True)
# evaluate means
df_gbp_means = model_eval.compute_and_plot_gbp(gbp_data,
current_feature_list,
True,
suppress_plots=True)
# group by feature type and rank by value
df_gbp_means = df_gbp_means.groupby('feature').mean().sort_values(
by='gbp', ascending=False).reset_index()
# get top ranking value and append to removal list
feature_to_remove = df_gbp_means.iloc[0, 0]
print("removing " + feature_to_remove + "...")
remove_features.append(feature_to_remove)
# create the mask
mask = np.isin(qty, remove_features, invert=True)
# mask the full state vector in config_dict global variable
config_dict['custom_state'] = full_state_index[mask]
current_feature_list = np.array(qty)[
mask] #update the current feature list
# decrease the input dimension of the model by one
input_state = input_state - 1
# redefine the models
print('redefining model with input state dims: {}'.format(input_state))
if model_type == "VS":
model = mdl.StateVisionModel(30,
input_state,
3,
feature_extract=feat_extract)
elif model_type == "S":
model = mdl.StateModel(input_state, 3)
# recalculate the number of parameters
param_count.append(count_params(model))
# redefine the optimizer
optimizer = opt.Adam(model.parameters(), lr=0.01)
# redefine the dataloader
dataloaders, dataset_sizes = dat.init_dataset(train_list, val_list,
val_list, model_type,
config_dict)
# retrain the model
model, train_history, val_history = mdl.train_model(
model,
criterion,
optimizer,
dataloaders,
dataset_sizes,
num_epochs=50,
model_type=model_type,
weight_file=weight_file,
no_pbar=True)
print('retraining completed')
# do inference
print('evaluating model predictions...')
predictions = mdl.evaluate_model(model,
dataloaders['test'],
model_type=model_type,
no_pbar=True)
# compute the loss statistics
print('computing performance metrics...')
metrics = model_eval.compute_loss_metrics(
predictions, dataloaders['test'].dataset.label_array[:, 1:4])
metrics_list.append(metrics)
print('Performance Summary:')
print(metrics)
return remove_features, param_count, metrics_list
# we need to unnormalize our predictions...
file_dir = '../experiment_data' # define the file directory for dataset
train_list = [1, 3, 5, 7, 8, 10, 12, 14, 15, 17, 19, 21]
val_list = [1]
config_dict = {
'file_dir': file_dir,
'include_torque': False,
'spatial_forces': False,
'custom_state': None,
'batch_size': 32,
'crop_list': [],
'trans_function': None
}
test_list = [1]
loader_dict, loader_sizes = dat.init_dataset(train_list, val_list, test_list,
'S', config_dict)
miu = loader_dict['train'].dataset.mean[1:4]
sigma = loader_dict['train'].dataset.stdev[1:4]
#%% TEST LIST SPEC
test_list_full = [
4, 11, 18, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 36, 37, 38, 39
]
condition_list = [
'center', 'right', 'left', 'right_less', 'right_less', 'right_more',
'right_more', 'left_less', 'left_less', 'left_more', 'left_more',
'new_tool', 'new_tool', 'new_material', 'new_material', 'center', 'right',
'left'
]