Commit 293569a4 authored by Christian Chapman-Bird's avatar Christian Chapman-Bird
Browse files

Dataloading for first batch of schwarz data updated

parent 9d2d3461

EMRI_DET/nn/model_train_test.py

+2 −2
Original line number Diff line number Diff line
@@ -54,7 +54,7 @@ def model_train_test(data, model, device, n_epochs, n_batches, loss_function, le 
                    for param in model.parameters():

                        param.grad = None

                    outputs = model(inputs[i * ytrainsize // n_batches:(i+1)*ytrainsize // n_batches])

                    loss = torch.sqrt(loss_function(outputs, targets[i * ytrainsize // n_batches: (i+1)*ytrainsize // n_batches]))

                    loss = loss_function(outputs, targets[i * ytrainsize // n_batches: (i+1)*ytrainsize // n_batches])

                    loss.backward()

                    optimizer.step()

                    current_loss += loss.item()
@@ -74,7 +74,7 @@ def model_train_test(data, model, device, n_epochs, n_batches, loss_function, le 


                    for i in range(n_batches):

                        outputs = model(inputs[i * ytestsize // n_batches: (i+1)*ytestsize // n_batches])

                        loss = torch.sqrt(loss_function(outputs, targets[i * ytestsize // n_batches: (i+1)*ytestsize // n_batches]))

                        loss = loss_function(outputs, targets[i * ytestsize // n_batches: (i+1)*ytestsize // n_batches])

                        current_loss += loss.item()



                    test_losses.append(current_loss / n_batches)

emri_data/schwarz_data/grid_dataframe.csv

0 → 100644
+0 −0

File added.

Preview suppressed by a .gitattributes entry or the file's encoding is unsupported.

View file

emri_data/schwarz_data/samp_dataframe.csv

0 → 100644
+0 −0

File added.

Preview suppressed by a .gitattributes entry or the file's encoding is unsupported.

View file

emri_data/scripts/data_preprocess_dataframe.py

0 → 100644
+62 −0
Original line number Diff line number Diff line 
import numpy as np

from pathlib import Path

import pandas as pd



data_directory = '../schwarz_data/{}'



# Grid

snr_grid = np.load(data_directory.format('grid_snrs.npy'))

intrinsics = np.load(data_directory.format('grid_intrinsics.npy'))

angulars = np.load(data_directory.format('grid_extrinsics.npy'))

plunges = np.load(data_directory.format('grid_plunges.npy'))



intrinsics[:,:2] = np.log10(intrinsics[:,:2])



grid_out = np.zeros(shape=(snr_grid.size,10))

flat_snr_grid = snr_grid.flatten()



num = 0

for intrinsic_set in intrinsics:

    for angular_set in angulars:

        for time in plunges:

            out = np.zeros(9)

            out[:5] = intrinsic_set

            out[5:8] = angular_set

            out[8] = time



            grid_out[num,:9] = out



            snr_here = flat_snr_grid[num]

            if snr_here == 0:

                snr_here += 1e-6

            grid_out[num,9] = snr_here

            num += 1



save_dir = '../schwarz_data/{}'

Path('../schwarz_data/').mkdir(parents=True, exist_ok=True)



cols = ['logM','logq','a','e','Y0','thetaS','phiS','thetaK','t','SNR']

df_out = pd.DataFrame(grid_out, columns=cols)

df_out.to_csv(data_directory.format('grid_dataframe.csv'), index=False)



# Samples

snr_list = np.load(data_directory.format('samp_snrs.npy'))

inds_to_keep = ~np.isnan(snr_list)



intrinsics = np.load(data_directory.format('samp_intrinsics.npy'))

angulars = np.load(data_directory.format('samp_extrinsics.npy'))

intrinsics[:,:2] = np.log10(intrinsics[:,:2])



samp_out = np.zeros(shape=(snr_list.size,10))



for i in range(snr_list.size):

    here = np.zeros(9)

    here[:5] = intrinsics[i,:5]

    here[5:8] = angulars[i,:]

    here[8] = intrinsics[i,5]

    samp_out[i,:9] = here

    samp_out[i,9] = snr_list[i]



samp_out = samp_out[inds_to_keep,:]

samp_df_out = pd.DataFrame(samp_out, columns=cols)

samp_df_out.to_csv(data_directory.format('samp_dataframe.csv'), index=False)