add possibility to convert time from seconds to hours

src/sparcscore/ml/datasets.py

@@ -210,6 +210,7 @@ class HDF5SingleCellDatasetRegression(Dataset):
  def __init__(self, 
  dir_list: list[str], 
  target_col: list[int],
+ hours: False,
  root_dir: str, 
  max_level: int = 5, 
  transform = None, 
@@ -219,6 +220,7 @@ def __init__(self,
 
  self.dir_list = dir_list # list of directories with hdf5 files
  self.target_col = target_col # list of indices for target columns, maps 1 to 1 with dir_list, i.e. target_col[i] is the target column for dir_list[i]
+ self.hours = hours # convert target to hours
  self.root_dir = root_dir 
  self.transform = transform 
  self.select_channel = select_channel
@@ -257,6 +259,8 @@ def add_hdf_to_index(self, path, target_col):
  self.handle_list.append(input_hdf.get('single_cell_data')) # append data handle (i.e. extracted images)
 
  for current_target, row in zip(current_target_col, index_handle): # iterate over rows in index handle, i.e. over all cells
+ if self.hours:
+ current_target = current_target / 3600 # convert seconds to hours
  self.data_locator.append([current_target, handle_id] + list(row)) # append target, handle id, and row to data locator
  except:
  return
@@ -281,14 +285,7 @@ def stats(self):
  targets = [info[0] for info in self.data_locator] # get all targets from data locator
  targets = np.array(targets, dtype=float) # convert to numpy array
 
- #min_target = np.min(targets)
- #max_target = np.max(targets)
-
- # add more stats eventually
-
  print(f"Total samples: {len(targets)}")
- #print(f"Min target: {min_target:.2f}")
- #print(f"Max target: {max_target:.2f}")
 
  def __len__(self):
  return len(self.data_locator) # return length of data locator
@@ -303,8 +300,6 @@ def __getitem__(self, idx):
  cell_tensor = self.handle_list[data_item[1]][data_item[2], self.select_channel] 
  t = torch.from_numpy(cell_tensor).float() # convert to float tensor
  t = torch.unsqueeze(t, 0) # add channel dimension to tensor
-
- #print(f"Selected channel {self.select_channel} from data.")
  else: 
  cell_tensor = self.handle_list[data_item[1]][data_item[2]] 
  t = torch.from_numpy(cell_tensor).float() # convert to float tensor

src/sparcscore/ml/plmodels.py

@@ -232,8 +232,6 @@ def test_step(self, batch):
  output = self.network(data)
  loss = F.huber_loss(output, target, delta=1.0, reduction='mean')
 
- # accuracy metrics for regression???
-
  self.log('loss/test', loss, on_step=False, on_epoch=True, prog_bar=True)
  self.log('mse/test', self.mse(output, target), on_epoch=True, prog_bar=True)
  self.log('mae/test', self.mae(output, target), on_epoch=True, prog_bar=True)