Tutorial:RNA+ADT (well-labeled)¶
In this tutorial, we will show how to cluster RNA+ADT(well-labeled) data using scMAGCA. As an example, we use a paired human peripheral blood sample dataset ‘10X1kpbmc’ containing 713 cells with two omics. Among them, ADT has 17 features, while RNA contains 33,538 features.
Loading package¶
[4]:
import numpy as np
import pandas as pd
import torch
import scanpy as sc
import random
import warnings
warnings.filterwarnings("ignore")
[5]:
from scMAGCA.preprocess import read_dataset, preprocess_dataset
from scMAGCA.utils import *
from scMAGCA.scMAGCA import scMultiCluster
[6]:
# set seed
random.seed(3407)
np.random.seed(3407)
torch.manual_seed(3407)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.enabled = False
torch.backends.cudnn.benchmark = False
Reading dataset¶
The required input files include:
x1: protein abundance matrix (data format is h5ad file) : 10X1kpbmc_adt.h5ad;
x2: Gene expression matrix (data format is h5ad file) : 10X1kpbmc_rna.h5ad;
Real label (stored as csv file) : 10X1kpbmc_label.csv.
To ensure reproducibility of the results, please read the above data as follows:
[7]:
x1 = np.array(sc.read_h5ad('../datasets/10X1kpbmc/10X1kpbmc_adt.h5ad').to_df()).astype('float32')
x2 = np.array(sc.read_h5ad('../datasets/10X1kpbmc/10X1kpbmc_rna.h5ad').to_df()).astype('float32')
y = np.array(pd.read_csv('../datasets/10X1kpbmc/10X1kpbmc_label.csv')['Cluster']).astype('float32')
[8]:
x1,x2,y
[8]:
(array([[3.000e+00, 6.000e+00, 2.000e+00, ..., 3.000e+00, 4.000e+00,
1.000e+00],
[3.444e+03, 3.995e+03, 4.900e+01, ..., 6.000e+00, 8.000e+00,
2.000e+00],
[1.733e+03, 3.401e+03, 2.300e+01, ..., 1.000e+00, 6.000e+00,
0.000e+00],
...,
[3.000e+00, 1.190e+02, 1.800e+01, ..., 1.000e+00, 1.000e+00,
2.000e+00],
[1.700e+01, 8.190e+02, 1.100e+01, ..., 3.000e+00, 2.000e+00,
2.000e+00],
[9.000e+00, 1.300e+01, 1.000e+01, ..., 4.000e+00, 1.100e+01,
5.000e+00]], dtype=float32),
array([[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
...,
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.]], dtype=float32),
array([5., 1., 1., 1., 1., 3., 3., 3., 1., 1., 3., 1., 1., 1., 5., 3., 3.,
3., 3., 2., 3., 1., 2., 1., 3., 4., 2., 5., 1., 1., 2., 5., 1., 4.,
2., 1., 3., 5., 2., 4., 1., 1., 3., 1., 1., 5., 1., 1., 1., 5., 5.,
2., 4., 1., 3., 3., 1., 4., 2., 5., 2., 1., 1., 1., 5., 1., 3., 2.,
3., 4., 2., 1., 2., 4., 3., 1., 3., 1., 1., 1., 1., 3., 4., 4., 2.,
4., 2., 3., 5., 5., 2., 4., 3., 4., 1., 1., 5., 2., 1., 1., 2., 1.,
1., 3., 1., 4., 1., 4., 1., 3., 4., 2., 5., 1., 1., 4., 1., 4., 2.,
5., 4., 4., 4., 2., 1., 5., 2., 1., 1., 3., 2., 2., 1., 1., 4., 3.,
4., 1., 4., 3., 1., 2., 1., 1., 5., 1., 3., 3., 3., 3., 5., 4., 3.,
4., 1., 1., 1., 2., 3., 3., 3., 2., 2., 1., 4., 3., 3., 4., 4., 4.,
1., 4., 1., 3., 5., 3., 5., 2., 3., 5., 1., 1., 2., 1., 3., 1., 2.,
1., 1., 5., 4., 4., 4., 2., 2., 5., 1., 2., 2., 2., 5., 2., 2., 1.,
2., 4., 3., 4., 1., 4., 3., 4., 1., 4., 3., 3., 2., 1., 2., 3., 4.,
3., 1., 2., 2., 2., 3., 4., 3., 5., 1., 2., 2., 3., 1., 2., 1., 4.,
5., 4., 2., 5., 1., 5., 2., 2., 3., 1., 1., 1., 1., 4., 5., 3., 4.,
3., 3., 4., 4., 1., 1., 1., 1., 4., 2., 1., 1., 1., 1., 3., 2., 2.,
1., 3., 2., 1., 4., 4., 4., 4., 2., 4., 3., 2., 1., 1., 5., 4., 1.,
2., 1., 2., 2., 1., 4., 1., 4., 1., 1., 3., 1., 4., 3., 1., 5., 2.,
3., 2., 1., 2., 2., 1., 4., 4., 1., 2., 2., 5., 1., 2., 1., 1., 3.,
2., 1., 1., 1., 4., 2., 2., 4., 4., 4., 4., 4., 4., 1., 2., 4., 3.,
4., 4., 2., 3., 4., 4., 1., 2., 1., 1., 1., 1., 5., 3., 2., 3., 1.,
1., 5., 3., 4., 1., 1., 3., 5., 2., 1., 5., 2., 1., 2., 3., 1., 1.,
1., 4., 3., 3., 2., 1., 3., 4., 2., 3., 1., 3., 1., 2., 2., 4., 1.,
3., 3., 4., 1., 1., 1., 1., 1., 5., 2., 1., 5., 5., 1., 3., 2., 1.,
1., 4., 4., 4., 3., 2., 5., 1., 3., 1., 1., 2., 2., 1., 1., 4., 3.,
1., 5., 5., 1., 4., 4., 1., 2., 1., 2., 2., 3., 1., 1., 2., 1., 1.,
5., 4., 5., 3., 4., 5., 1., 2., 1., 2., 5., 4., 3., 3., 1., 1., 3.,
1., 2., 5., 1., 3., 2., 5., 1., 3., 2., 1., 2., 5., 2., 1., 2., 1.,
1., 2., 3., 1., 2., 3., 2., 2., 2., 5., 1., 4., 5., 1., 3., 4., 5.,
4., 2., 1., 1., 1., 2., 5., 2., 3., 4., 4., 5., 1., 2., 1., 2., 1.,
4., 5., 5., 1., 3., 5., 3., 1., 2., 1., 4., 1., 1., 5., 5., 2., 4.,
2., 1., 4., 2., 3., 4., 3., 4., 3., 2., 5., 4., 1., 3., 3., 3., 2.,
2., 3., 1., 1., 1., 1., 2., 1., 4., 2., 3., 3., 1., 2., 4., 2., 4.,
2., 1., 4., 4., 1., 1., 2., 5., 3., 1., 1., 3., 2., 4., 1., 4., 2.,
1., 4., 4., 2., 4., 1., 4., 4., 1., 5., 4., 2., 3., 1., 5., 1., 4.,
2., 1., 1., 1., 3., 5., 3., 2., 3., 1., 1., 2., 5., 4., 1., 3., 4.,
4., 1., 2., 4., 2., 5., 1., 3., 3., 3., 3., 2., 1., 4., 2., 2., 3.,
5., 3., 1., 1., 3., 5., 2., 5., 4., 3., 4., 1., 2., 1., 2., 5., 2.,
5., 3., 1., 3., 1., 1., 1., 2., 4., 2., 1., 4., 2., 2., 3., 1., 3.,
1., 3., 2., 1., 3., 5., 5., 4., 2., 3., 2., 5., 3., 1., 2., 4., 1.,
3., 2., 3., 1., 4., 1., 4., 2., 2., 2., 1., 1., 2., 5., 1., 1., 2.,
2., 1., 5., 5., 5., 1., 3., 2., 1., 1., 4., 1., 2., 4., 3., 5.],
dtype=float32))
Due to the small number of features in ADT omics data and the large gap between the feature dimensions of RNA omics, for RNA+ADT data, we only select high-expression features for RNA omics (the default number of chosen genes is 2000).
[11]:
importantGenes = geneSelection(x2, n=2000)
x2 = x2[:, importantGenes]
Chosen offset: 0.21
[12]:
adata1 = sc.AnnData(x1)
adata1 = read_dataset(adata1, copy=True)
adata1 = preprocess_dataset(adata1, normalize_input=True, logtrans_input=True)
### Autoencoder: Successfully preprocessed 17 features and 713 cells.
[13]:
adata1
[13]:
AnnData object with n_obs × n_vars = 713 × 17
obs: 'DCA_split', 'size_factors'
var: 'mean', 'std'
uns: 'log1p'
[14]:
adata2 = sc.AnnData(x2)
adata2 = read_dataset(adata2, copy=True)
adata2 = preprocess_dataset(adata2, normalize_input=True, logtrans_input=True)
### Autoencoder: Successfully preprocessed 1999 features and 713 cells.
[15]:
adata2
[15]:
AnnData object with n_obs × n_vars = 713 × 1999
obs: 'DCA_split', 'size_factors'
var: 'mean', 'std'
uns: 'log1p'
Training the model¶
Our model training is divided into two stages: pre-training stage and formal training stage. The number of epochs in the pre-training phase is set to 400, while the number in the formal training is 2000. In formal training, if the change in training results reaches a set threshold, it will end early.
After the training, the potential representations and predicted label values obtained by scMAGCA are saved to the corresponding folder of the data set and can be used for subsequent downstream analysis.
[16]:
model = scMultiCluster(input_dim1=adata1.n_vars,input_dim2=adata2.n_vars,
alpha=0.5,beta=1,gama=0.1,device='cuda').to('cuda')
[17]:
model
[17]:
scMultiCluster(
(encoder): Encoder(
(stacked_gnn): ModuleList(
(0): GCNConv(2016, 1024)
(1): GCNConv(1024, 256)
(2): GCNConv(256, 64)
(3): GCNConv(64, 32)
)
(stacked_bns): ModuleList(
(0): BatchNorm1d(1024, eps=1e-05, momentum=0.01, affine=True, track_running_stats=True)
(1): BatchNorm1d(256, eps=1e-05, momentum=0.01, affine=True, track_running_stats=True)
(2): BatchNorm1d(64, eps=1e-05, momentum=0.01, affine=True, track_running_stats=True)
(3): BatchNorm1d(32, eps=1e-05, momentum=0.01, affine=True, track_running_stats=True)
)
(stacked_prelus): ModuleList(
(0-3): 4 x PReLU(num_parameters=1)
)
)
(decoder): Sequential(
(0): Linear(in_features=32, out_features=512, bias=True)
(1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(2): PReLU(num_parameters=1)
(3): Linear(in_features=512, out_features=1024, bias=True)
(4): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(5): PReLU(num_parameters=1)
(6): Linear(in_features=1024, out_features=2016, bias=True)
)
(dec_mean): Sequential(
(0): Linear(in_features=32, out_features=256, bias=True)
(1): Linear(in_features=256, out_features=512, bias=True)
(2): Linear(in_features=512, out_features=2016, bias=True)
(3): MeanAct()
)
(dec_disp): Sequential(
(0): Linear(in_features=32, out_features=256, bias=True)
(1): Linear(in_features=256, out_features=512, bias=True)
(2): Linear(in_features=512, out_features=2016, bias=True)
(3): DispAct()
)
(dec_pi): Sequential(
(0): Linear(in_features=32, out_features=256, bias=True)
(1): Linear(in_features=256, out_features=512, bias=True)
(2): Linear(in_features=512, out_features=2016, bias=True)
(3): Sigmoid()
)
(zinb_loss): ZINBLoss()
)
[18]:
pretrain_latent = model.pretrain_autoencoder(
X1=adata1.X, X2=adata2.X, X1_raw=adata1.raw.X, X2_raw=adata2.raw.X,
epochs=400, file='10X1kpbmc')
Pretraining stage
Pretrain epoch 1, recon_loss:1.179670, zinb_loss:3.310244, adversial_loss:1.358601
Pretrain epoch 2, recon_loss:1.053315, zinb_loss:2.773893, adversial_loss:1.373735
Pretrain epoch 3, recon_loss:0.862565, zinb_loss:2.396230, adversial_loss:1.360926
Pretrain epoch 4, recon_loss:0.777968, zinb_loss:2.083947, adversial_loss:1.360451
Pretrain epoch 5, recon_loss:0.754381, zinb_loss:1.846060, adversial_loss:1.360216
Pretrain epoch 6, recon_loss:0.733581, zinb_loss:1.646314, adversial_loss:1.355564
Pretrain epoch 7, recon_loss:0.721840, zinb_loss:1.484166, adversial_loss:1.350406
Pretrain epoch 8, recon_loss:0.717826, zinb_loss:1.361452, adversial_loss:1.347699
Pretrain epoch 9, recon_loss:0.708908, zinb_loss:1.277065, adversial_loss:1.346102
Pretrain epoch 10, recon_loss:0.694630, zinb_loss:1.224692, adversial_loss:1.344238
Pretrain epoch 11, recon_loss:0.684225, zinb_loss:1.194159, adversial_loss:1.342163
Pretrain epoch 12, recon_loss:0.680018, zinb_loss:1.175642, adversial_loss:1.340560
Pretrain epoch 13, recon_loss:0.677840, zinb_loss:1.162826, adversial_loss:1.339098
Pretrain epoch 14, recon_loss:0.675129, zinb_loss:1.152876, adversial_loss:1.337435
Pretrain epoch 15, recon_loss:0.672355, zinb_loss:1.144369, adversial_loss:1.335139
Pretrain epoch 16, recon_loss:0.670228, zinb_loss:1.136601, adversial_loss:1.332293
Pretrain epoch 17, recon_loss:0.667790, zinb_loss:1.129140, adversial_loss:1.329263
Pretrain epoch 18, recon_loss:0.664680, zinb_loss:1.121876, adversial_loss:1.326783
Pretrain epoch 19, recon_loss:0.662127, zinb_loss:1.115053, adversial_loss:1.324961
Pretrain epoch 20, recon_loss:0.660377, zinb_loss:1.108180, adversial_loss:1.323852
Pretrain epoch 21, recon_loss:0.659001, zinb_loss:1.100873, adversial_loss:1.323825
Pretrain epoch 22, recon_loss:0.656459, zinb_loss:1.093486, adversial_loss:1.322779
Pretrain epoch 23, recon_loss:0.654088, zinb_loss:1.086390, adversial_loss:1.321610
Pretrain epoch 24, recon_loss:0.652294, zinb_loss:1.079503, adversial_loss:1.320213
Pretrain epoch 25, recon_loss:0.650671, zinb_loss:1.072774, adversial_loss:1.318725
Pretrain epoch 26, recon_loss:0.648966, zinb_loss:1.066303, adversial_loss:1.317523
Pretrain epoch 27, recon_loss:0.646991, zinb_loss:1.060190, adversial_loss:1.316257
Pretrain epoch 28, recon_loss:0.645398, zinb_loss:1.054652, adversial_loss:1.314740
Pretrain epoch 29, recon_loss:0.643983, zinb_loss:1.049268, adversial_loss:1.312217
Pretrain epoch 30, recon_loss:0.643025, zinb_loss:1.043892, adversial_loss:1.308721
Pretrain epoch 31, recon_loss:0.642111, zinb_loss:1.039183, adversial_loss:1.306067
Pretrain epoch 32, recon_loss:0.640455, zinb_loss:1.034951, adversial_loss:1.304514
Pretrain epoch 33, recon_loss:0.639247, zinb_loss:1.031011, adversial_loss:1.303380
Pretrain epoch 34, recon_loss:0.637743, zinb_loss:1.027282, adversial_loss:1.301915
Pretrain epoch 35, recon_loss:0.636664, zinb_loss:1.023664, adversial_loss:1.300623
Pretrain epoch 36, recon_loss:0.635682, zinb_loss:1.020120, adversial_loss:1.299578
Pretrain epoch 37, recon_loss:0.634332, zinb_loss:1.016628, adversial_loss:1.298586
Pretrain epoch 38, recon_loss:0.633481, zinb_loss:1.013218, adversial_loss:1.298119
Pretrain epoch 39, recon_loss:0.632329, zinb_loss:1.010160, adversial_loss:1.296602
Pretrain epoch 40, recon_loss:0.631456, zinb_loss:1.007311, adversial_loss:1.295990
Pretrain epoch 41, recon_loss:0.630563, zinb_loss:1.004686, adversial_loss:1.294802
Pretrain epoch 42, recon_loss:0.629735, zinb_loss:1.002095, adversial_loss:1.295299
Pretrain epoch 43, recon_loss:0.629020, zinb_loss:1.000048, adversial_loss:1.293201
Pretrain epoch 44, recon_loss:0.628373, zinb_loss:0.997680, adversial_loss:1.293849
Pretrain epoch 45, recon_loss:0.627905, zinb_loss:0.995719, adversial_loss:1.290645
Pretrain epoch 46, recon_loss:0.626810, zinb_loss:0.993385, adversial_loss:1.290701
Pretrain epoch 47, recon_loss:0.626094, zinb_loss:0.991341, adversial_loss:1.289930
Pretrain epoch 48, recon_loss:0.625388, zinb_loss:0.989542, adversial_loss:1.288692
Pretrain epoch 49, recon_loss:0.624706, zinb_loss:0.987548, adversial_loss:1.288829
Pretrain epoch 50, recon_loss:0.624136, zinb_loss:0.986056, adversial_loss:1.287461
Pretrain epoch 51, recon_loss:0.623274, zinb_loss:0.984164, adversial_loss:1.287607
Pretrain epoch 52, recon_loss:0.622424, zinb_loss:0.982464, adversial_loss:1.286898
Pretrain epoch 53, recon_loss:0.621879, zinb_loss:0.981260, adversial_loss:1.285860
Pretrain epoch 54, recon_loss:0.621108, zinb_loss:0.979639, adversial_loss:1.285778
Pretrain epoch 55, recon_loss:0.620530, zinb_loss:0.978340, adversial_loss:1.284826
Pretrain epoch 56, recon_loss:0.619800, zinb_loss:0.977073, adversial_loss:1.284936
Pretrain epoch 57, recon_loss:0.619347, zinb_loss:0.975783, adversial_loss:1.283673
Pretrain epoch 58, recon_loss:0.618823, zinb_loss:0.974526, adversial_loss:1.284123
Pretrain epoch 59, recon_loss:0.618686, zinb_loss:0.973556, adversial_loss:1.282676
Pretrain epoch 60, recon_loss:0.617441, zinb_loss:0.972172, adversial_loss:1.283096
Pretrain epoch 61, recon_loss:0.616835, zinb_loss:0.970980, adversial_loss:1.281648
Pretrain epoch 62, recon_loss:0.615886, zinb_loss:0.969648, adversial_loss:1.281288
Pretrain epoch 63, recon_loss:0.615350, zinb_loss:0.968580, adversial_loss:1.281277
Pretrain epoch 64, recon_loss:0.614922, zinb_loss:0.967744, adversial_loss:1.280465
Pretrain epoch 65, recon_loss:0.614932, zinb_loss:0.967338, adversial_loss:1.280521
Pretrain epoch 66, recon_loss:0.614976, zinb_loss:0.966357, adversial_loss:1.280103
Pretrain epoch 67, recon_loss:0.613583, zinb_loss:0.965226, adversial_loss:1.280383
Pretrain epoch 68, recon_loss:0.613921, zinb_loss:0.965229, adversial_loss:1.278780
Pretrain epoch 69, recon_loss:0.612545, zinb_loss:0.963646, adversial_loss:1.279052
Pretrain epoch 70, recon_loss:0.612793, zinb_loss:0.963037, adversial_loss:1.279510
Pretrain epoch 71, recon_loss:0.611815, zinb_loss:0.961470, adversial_loss:1.278643
Pretrain epoch 72, recon_loss:0.611546, zinb_loss:0.960857, adversial_loss:1.278122
Pretrain epoch 73, recon_loss:0.610706, zinb_loss:0.959825, adversial_loss:1.278297
Pretrain epoch 74, recon_loss:0.610377, zinb_loss:0.959457, adversial_loss:1.278214
Pretrain epoch 75, recon_loss:0.608972, zinb_loss:0.958782, adversial_loss:1.277103
Pretrain epoch 76, recon_loss:0.609311, zinb_loss:0.957827, adversial_loss:1.276618
Pretrain epoch 77, recon_loss:0.608211, zinb_loss:0.957331, adversial_loss:1.277107
Pretrain epoch 78, recon_loss:0.607739, zinb_loss:0.956897, adversial_loss:1.276694
Pretrain epoch 79, recon_loss:0.607406, zinb_loss:0.956075, adversial_loss:1.276365
Pretrain epoch 80, recon_loss:0.607084, zinb_loss:0.955484, adversial_loss:1.276402
Pretrain epoch 81, recon_loss:0.605922, zinb_loss:0.954734, adversial_loss:1.275925
Pretrain epoch 82, recon_loss:0.605686, zinb_loss:0.954194, adversial_loss:1.275973
Pretrain epoch 83, recon_loss:0.605507, zinb_loss:0.954741, adversial_loss:1.275807
Pretrain epoch 84, recon_loss:0.606017, zinb_loss:0.955828, adversial_loss:1.275457
Pretrain epoch 85, recon_loss:0.606689, zinb_loss:0.955475, adversial_loss:1.274885
Pretrain epoch 86, recon_loss:0.603597, zinb_loss:0.952265, adversial_loss:1.275342
Pretrain epoch 87, recon_loss:0.603961, zinb_loss:0.952641, adversial_loss:1.275476
Pretrain epoch 88, recon_loss:0.603907, zinb_loss:0.952613, adversial_loss:1.274445
Pretrain epoch 89, recon_loss:0.602533, zinb_loss:0.950886, adversial_loss:1.273964
Pretrain epoch 90, recon_loss:0.601846, zinb_loss:0.951025, adversial_loss:1.274571
Pretrain epoch 91, recon_loss:0.601666, zinb_loss:0.950418, adversial_loss:1.274107
Pretrain epoch 92, recon_loss:0.600451, zinb_loss:0.950003, adversial_loss:1.272773
Pretrain epoch 93, recon_loss:0.600263, zinb_loss:0.949291, adversial_loss:1.273564
Pretrain epoch 94, recon_loss:0.599480, zinb_loss:0.948868, adversial_loss:1.273637
Pretrain epoch 95, recon_loss:0.599118, zinb_loss:0.948717, adversial_loss:1.272256
Pretrain epoch 96, recon_loss:0.598330, zinb_loss:0.947870, adversial_loss:1.272520
Pretrain epoch 97, recon_loss:0.598428, zinb_loss:0.948256, adversial_loss:1.272994
Pretrain epoch 98, recon_loss:0.599205, zinb_loss:0.948368, adversial_loss:1.272520
Pretrain epoch 99, recon_loss:0.600360, zinb_loss:0.950071, adversial_loss:1.271988
Pretrain epoch 100, recon_loss:0.597603, zinb_loss:0.946913, adversial_loss:1.272954
Pretrain epoch 101, recon_loss:0.598265, zinb_loss:0.948654, adversial_loss:1.273010
Pretrain epoch 102, recon_loss:0.598125, zinb_loss:0.946708, adversial_loss:1.272140
Pretrain epoch 103, recon_loss:0.596699, zinb_loss:0.946644, adversial_loss:1.271546
Pretrain epoch 104, recon_loss:0.596948, zinb_loss:0.945447, adversial_loss:1.271937
Pretrain epoch 105, recon_loss:0.594884, zinb_loss:0.945366, adversial_loss:1.271829
Pretrain epoch 106, recon_loss:0.595728, zinb_loss:0.944949, adversial_loss:1.270860
Pretrain epoch 107, recon_loss:0.593774, zinb_loss:0.944627, adversial_loss:1.270213
Pretrain epoch 108, recon_loss:0.593907, zinb_loss:0.944016, adversial_loss:1.270924
Pretrain epoch 109, recon_loss:0.592436, zinb_loss:0.943547, adversial_loss:1.270654
Pretrain epoch 110, recon_loss:0.593074, zinb_loss:0.943505, adversial_loss:1.269729
Pretrain epoch 111, recon_loss:0.591259, zinb_loss:0.942827, adversial_loss:1.269440
Pretrain epoch 112, recon_loss:0.591416, zinb_loss:0.943134, adversial_loss:1.269742
Pretrain epoch 113, recon_loss:0.590118, zinb_loss:0.942558, adversial_loss:1.268840
Pretrain epoch 114, recon_loss:0.591088, zinb_loss:0.942681, adversial_loss:1.268550
Pretrain epoch 115, recon_loss:0.591063, zinb_loss:0.943514, adversial_loss:1.268707
Pretrain epoch 116, recon_loss:0.591478, zinb_loss:0.946264, adversial_loss:1.268885
Pretrain epoch 117, recon_loss:0.592222, zinb_loss:0.947532, adversial_loss:1.267890
Pretrain epoch 118, recon_loss:0.590538, zinb_loss:0.942911, adversial_loss:1.267889
Pretrain epoch 119, recon_loss:0.588685, zinb_loss:0.942405, adversial_loss:1.267961
Pretrain epoch 120, recon_loss:0.590820, zinb_loss:0.943007, adversial_loss:1.267911
Pretrain epoch 121, recon_loss:0.587976, zinb_loss:0.941211, adversial_loss:1.267495
Pretrain epoch 122, recon_loss:0.588107, zinb_loss:0.941682, adversial_loss:1.267016
Pretrain epoch 123, recon_loss:0.587850, zinb_loss:0.941220, adversial_loss:1.267151
Pretrain epoch 124, recon_loss:0.586103, zinb_loss:0.940096, adversial_loss:1.266919
Pretrain epoch 125, recon_loss:0.586695, zinb_loss:0.940915, adversial_loss:1.266283
Pretrain epoch 126, recon_loss:0.584600, zinb_loss:0.939462, adversial_loss:1.266199
Pretrain epoch 127, recon_loss:0.584715, zinb_loss:0.939992, adversial_loss:1.266392
Pretrain epoch 128, recon_loss:0.583646, zinb_loss:0.939460, adversial_loss:1.265532
Pretrain epoch 129, recon_loss:0.583115, zinb_loss:0.938871, adversial_loss:1.265500
Pretrain epoch 130, recon_loss:0.582615, zinb_loss:0.939027, adversial_loss:1.265729
Pretrain epoch 131, recon_loss:0.581923, zinb_loss:0.938283, adversial_loss:1.265340
Pretrain epoch 132, recon_loss:0.581585, zinb_loss:0.938320, adversial_loss:1.264889
Pretrain epoch 133, recon_loss:0.580901, zinb_loss:0.937896, adversial_loss:1.265080
Pretrain epoch 134, recon_loss:0.580363, zinb_loss:0.937661, adversial_loss:1.264922
Pretrain epoch 135, recon_loss:0.579973, zinb_loss:0.937575, adversial_loss:1.264256
Pretrain epoch 136, recon_loss:0.579230, zinb_loss:0.937157, adversial_loss:1.264609
Pretrain epoch 137, recon_loss:0.578799, zinb_loss:0.937015, adversial_loss:1.264434
Pretrain epoch 138, recon_loss:0.578477, zinb_loss:0.936980, adversial_loss:1.263689
Pretrain epoch 139, recon_loss:0.578071, zinb_loss:0.936623, adversial_loss:1.264085
Pretrain epoch 140, recon_loss:0.577605, zinb_loss:0.936654, adversial_loss:1.263607
Pretrain epoch 141, recon_loss:0.577493, zinb_loss:0.936722, adversial_loss:1.263513
Pretrain epoch 142, recon_loss:0.577291, zinb_loss:0.937000, adversial_loss:1.263335
Pretrain epoch 143, recon_loss:0.576545, zinb_loss:0.937728, adversial_loss:1.263499
Pretrain epoch 144, recon_loss:0.578736, zinb_loss:0.939613, adversial_loss:1.262442
Pretrain epoch 145, recon_loss:0.582711, zinb_loss:0.939815, adversial_loss:1.263414
Pretrain epoch 146, recon_loss:0.582419, zinb_loss:0.936777, adversial_loss:1.262378
Pretrain epoch 147, recon_loss:0.577070, zinb_loss:0.935723, adversial_loss:1.262114
Pretrain epoch 148, recon_loss:0.577505, zinb_loss:0.937018, adversial_loss:1.262499
Pretrain epoch 149, recon_loss:0.577640, zinb_loss:0.935947, adversial_loss:1.261826
Pretrain epoch 150, recon_loss:0.575486, zinb_loss:0.934979, adversial_loss:1.261736
Pretrain epoch 151, recon_loss:0.575607, zinb_loss:0.935294, adversial_loss:1.262205
Pretrain epoch 152, recon_loss:0.574304, zinb_loss:0.935022, adversial_loss:1.261621
Pretrain epoch 153, recon_loss:0.575306, zinb_loss:0.934522, adversial_loss:1.261178
Pretrain epoch 154, recon_loss:0.572280, zinb_loss:0.934396, adversial_loss:1.261574
Pretrain epoch 155, recon_loss:0.572240, zinb_loss:0.934089, adversial_loss:1.261276
Pretrain epoch 156, recon_loss:0.571514, zinb_loss:0.934011, adversial_loss:1.261041
Pretrain epoch 157, recon_loss:0.570485, zinb_loss:0.933899, adversial_loss:1.260774
Pretrain epoch 158, recon_loss:0.569664, zinb_loss:0.933655, adversial_loss:1.260707
Pretrain epoch 159, recon_loss:0.569227, zinb_loss:0.933532, adversial_loss:1.260713
Pretrain epoch 160, recon_loss:0.568005, zinb_loss:0.933317, adversial_loss:1.260352
Pretrain epoch 161, recon_loss:0.568140, zinb_loss:0.933409, adversial_loss:1.260047
Pretrain epoch 162, recon_loss:0.568237, zinb_loss:0.933763, adversial_loss:1.260389
Pretrain epoch 163, recon_loss:0.567785, zinb_loss:0.935789, adversial_loss:1.259585
Pretrain epoch 164, recon_loss:0.572782, zinb_loss:0.941099, adversial_loss:1.260909
Pretrain epoch 165, recon_loss:0.574854, zinb_loss:0.941256, adversial_loss:1.260053
Pretrain epoch 166, recon_loss:0.571536, zinb_loss:0.934069, adversial_loss:1.259973
Pretrain epoch 167, recon_loss:0.574416, zinb_loss:0.936030, adversial_loss:1.260518
Pretrain epoch 168, recon_loss:0.569643, zinb_loss:0.934969, adversial_loss:1.259938
Pretrain epoch 169, recon_loss:0.569789, zinb_loss:0.933940, adversial_loss:1.259906
Pretrain epoch 170, recon_loss:0.567910, zinb_loss:0.933777, adversial_loss:1.260010
Pretrain epoch 171, recon_loss:0.568514, zinb_loss:0.933285, adversial_loss:1.259891
Pretrain epoch 172, recon_loss:0.565946, zinb_loss:0.933049, adversial_loss:1.259461
Pretrain epoch 173, recon_loss:0.566411, zinb_loss:0.932422, adversial_loss:1.259835
Pretrain epoch 174, recon_loss:0.563918, zinb_loss:0.932656, adversial_loss:1.259749
Pretrain epoch 175, recon_loss:0.564549, zinb_loss:0.932068, adversial_loss:1.259296
Pretrain epoch 176, recon_loss:0.562717, zinb_loss:0.931993, adversial_loss:1.259141
Pretrain epoch 177, recon_loss:0.562190, zinb_loss:0.931494, adversial_loss:1.259244
Pretrain epoch 178, recon_loss:0.561238, zinb_loss:0.931572, adversial_loss:1.259254
Pretrain epoch 179, recon_loss:0.560794, zinb_loss:0.931173, adversial_loss:1.258906
Pretrain epoch 180, recon_loss:0.559483, zinb_loss:0.931114, adversial_loss:1.258742
Pretrain epoch 181, recon_loss:0.558810, zinb_loss:0.930846, adversial_loss:1.258673
Pretrain epoch 182, recon_loss:0.557942, zinb_loss:0.930764, adversial_loss:1.258663
Pretrain epoch 183, recon_loss:0.557357, zinb_loss:0.930695, adversial_loss:1.258276
Pretrain epoch 184, recon_loss:0.556984, zinb_loss:0.930571, adversial_loss:1.258297
Pretrain epoch 185, recon_loss:0.556746, zinb_loss:0.930732, adversial_loss:1.257903
Pretrain epoch 186, recon_loss:0.558162, zinb_loss:0.931001, adversial_loss:1.258457
Pretrain epoch 187, recon_loss:0.563326, zinb_loss:0.931754, adversial_loss:1.257512
Pretrain epoch 188, recon_loss:0.562017, zinb_loss:0.931675, adversial_loss:1.258373
Pretrain epoch 189, recon_loss:0.558887, zinb_loss:0.931083, adversial_loss:1.256987
Pretrain epoch 190, recon_loss:0.555452, zinb_loss:0.930072, adversial_loss:1.257520
Pretrain epoch 191, recon_loss:0.556445, zinb_loss:0.930590, adversial_loss:1.257479
Pretrain epoch 192, recon_loss:0.555254, zinb_loss:0.931529, adversial_loss:1.257116
Pretrain epoch 193, recon_loss:0.555493, zinb_loss:0.930609, adversial_loss:1.257220
Pretrain epoch 194, recon_loss:0.554808, zinb_loss:0.929856, adversial_loss:1.256895
Pretrain epoch 195, recon_loss:0.554252, zinb_loss:0.929679, adversial_loss:1.256472
Pretrain epoch 196, recon_loss:0.553550, zinb_loss:0.929222, adversial_loss:1.257241
Pretrain epoch 197, recon_loss:0.550400, zinb_loss:0.928963, adversial_loss:1.256553
Pretrain epoch 198, recon_loss:0.550799, zinb_loss:0.929040, adversial_loss:1.256077
Pretrain epoch 199, recon_loss:0.550205, zinb_loss:0.928932, adversial_loss:1.256700
Pretrain epoch 200, recon_loss:0.548488, zinb_loss:0.928861, adversial_loss:1.256489
Pretrain epoch 201, recon_loss:0.549529, zinb_loss:0.929483, adversial_loss:1.255842
Pretrain epoch 202, recon_loss:0.552430, zinb_loss:0.930490, adversial_loss:1.255978
Pretrain epoch 203, recon_loss:0.555417, zinb_loss:0.932157, adversial_loss:1.256537
Pretrain epoch 204, recon_loss:0.552969, zinb_loss:0.932155, adversial_loss:1.255958
Pretrain epoch 205, recon_loss:0.548828, zinb_loss:0.928827, adversial_loss:1.255473
Pretrain epoch 206, recon_loss:0.547778, zinb_loss:0.928655, adversial_loss:1.255564
Pretrain epoch 207, recon_loss:0.549267, zinb_loss:0.929921, adversial_loss:1.255289
Pretrain epoch 208, recon_loss:0.547404, zinb_loss:0.929111, adversial_loss:1.255101
Pretrain epoch 209, recon_loss:0.545754, zinb_loss:0.928058, adversial_loss:1.255139
Pretrain epoch 210, recon_loss:0.545914, zinb_loss:0.928491, adversial_loss:1.255020
Pretrain epoch 211, recon_loss:0.544403, zinb_loss:0.928265, adversial_loss:1.254906
Pretrain epoch 212, recon_loss:0.543415, zinb_loss:0.927735, adversial_loss:1.255028
Pretrain epoch 213, recon_loss:0.543553, zinb_loss:0.927718, adversial_loss:1.254846
Pretrain epoch 214, recon_loss:0.540635, zinb_loss:0.927378, adversial_loss:1.254630
Pretrain epoch 215, recon_loss:0.541757, zinb_loss:0.927329, adversial_loss:1.254627
Pretrain epoch 216, recon_loss:0.540345, zinb_loss:0.927143, adversial_loss:1.254506
Pretrain epoch 217, recon_loss:0.538196, zinb_loss:0.926944, adversial_loss:1.254498
Pretrain epoch 218, recon_loss:0.538600, zinb_loss:0.926930, adversial_loss:1.254115
Pretrain epoch 219, recon_loss:0.537008, zinb_loss:0.926748, adversial_loss:1.254303
Pretrain epoch 220, recon_loss:0.536594, zinb_loss:0.926706, adversial_loss:1.254115
Pretrain epoch 221, recon_loss:0.539298, zinb_loss:0.926827, adversial_loss:1.253990
Pretrain epoch 222, recon_loss:0.545875, zinb_loss:0.927305, adversial_loss:1.253899
Pretrain epoch 223, recon_loss:0.547195, zinb_loss:0.927331, adversial_loss:1.254571
Pretrain epoch 224, recon_loss:0.542174, zinb_loss:0.927121, adversial_loss:1.253541
Pretrain epoch 225, recon_loss:0.541037, zinb_loss:0.926618, adversial_loss:1.253116
Pretrain epoch 226, recon_loss:0.541850, zinb_loss:0.926372, adversial_loss:1.253573
Pretrain epoch 227, recon_loss:0.537266, zinb_loss:0.926366, adversial_loss:1.253625
Pretrain epoch 228, recon_loss:0.538527, zinb_loss:0.926530, adversial_loss:1.253038
Pretrain epoch 229, recon_loss:0.536627, zinb_loss:0.926048, adversial_loss:1.253560
Pretrain epoch 230, recon_loss:0.534582, zinb_loss:0.925871, adversial_loss:1.253318
Pretrain epoch 231, recon_loss:0.534135, zinb_loss:0.925740, adversial_loss:1.253361
Pretrain epoch 232, recon_loss:0.532720, zinb_loss:0.925627, adversial_loss:1.253266
Pretrain epoch 233, recon_loss:0.531950, zinb_loss:0.925653, adversial_loss:1.252956
Pretrain epoch 234, recon_loss:0.531895, zinb_loss:0.925733, adversial_loss:1.253239
Pretrain epoch 235, recon_loss:0.532459, zinb_loss:0.926130, adversial_loss:1.252955
Pretrain epoch 236, recon_loss:0.536625, zinb_loss:0.927588, adversial_loss:1.252945
Pretrain epoch 237, recon_loss:0.539024, zinb_loss:0.930635, adversial_loss:1.253257
Pretrain epoch 238, recon_loss:0.537922, zinb_loss:0.931437, adversial_loss:1.252468
Pretrain epoch 239, recon_loss:0.532929, zinb_loss:0.928103, adversial_loss:1.252972
Pretrain epoch 240, recon_loss:0.535495, zinb_loss:0.926489, adversial_loss:1.252351
Pretrain epoch 241, recon_loss:0.533118, zinb_loss:0.926977, adversial_loss:1.252387
Pretrain epoch 242, recon_loss:0.532192, zinb_loss:0.926168, adversial_loss:1.252556
Pretrain epoch 243, recon_loss:0.531266, zinb_loss:0.925452, adversial_loss:1.252050
Pretrain epoch 244, recon_loss:0.529536, zinb_loss:0.925518, adversial_loss:1.252217
Pretrain epoch 245, recon_loss:0.529919, zinb_loss:0.925040, adversial_loss:1.252160
Pretrain epoch 246, recon_loss:0.526950, zinb_loss:0.924811, adversial_loss:1.252068
Pretrain epoch 247, recon_loss:0.526884, zinb_loss:0.924734, adversial_loss:1.252079
Pretrain epoch 248, recon_loss:0.524632, zinb_loss:0.924482, adversial_loss:1.252126
Pretrain epoch 249, recon_loss:0.523845, zinb_loss:0.924345, adversial_loss:1.251854
Pretrain epoch 250, recon_loss:0.523529, zinb_loss:0.924314, adversial_loss:1.252120
Pretrain epoch 251, recon_loss:0.521685, zinb_loss:0.924291, adversial_loss:1.251458
Pretrain epoch 252, recon_loss:0.520432, zinb_loss:0.924136, adversial_loss:1.251710
Pretrain epoch 253, recon_loss:0.520313, zinb_loss:0.924272, adversial_loss:1.251487
Pretrain epoch 254, recon_loss:0.521664, zinb_loss:0.924356, adversial_loss:1.251635
Pretrain epoch 255, recon_loss:0.523035, zinb_loss:0.924369, adversial_loss:1.251112
Pretrain epoch 256, recon_loss:0.525831, zinb_loss:0.924372, adversial_loss:1.251610
Pretrain epoch 257, recon_loss:0.522973, zinb_loss:0.924103, adversial_loss:1.250789
Pretrain epoch 258, recon_loss:0.518241, zinb_loss:0.923893, adversial_loss:1.251200
Pretrain epoch 259, recon_loss:0.517817, zinb_loss:0.923991, adversial_loss:1.250929
Pretrain epoch 260, recon_loss:0.518039, zinb_loss:0.923839, adversial_loss:1.250690
Pretrain epoch 261, recon_loss:0.515956, zinb_loss:0.923447, adversial_loss:1.250928
Pretrain epoch 262, recon_loss:0.513895, zinb_loss:0.923440, adversial_loss:1.250620
Pretrain epoch 263, recon_loss:0.513835, zinb_loss:0.923419, adversial_loss:1.250605
Pretrain epoch 264, recon_loss:0.512982, zinb_loss:0.923185, adversial_loss:1.250534
Pretrain epoch 265, recon_loss:0.512673, zinb_loss:0.923089, adversial_loss:1.250627
Pretrain epoch 266, recon_loss:0.512133, zinb_loss:0.923116, adversial_loss:1.250002
Pretrain epoch 267, recon_loss:0.511962, zinb_loss:0.923176, adversial_loss:1.250749
Pretrain epoch 268, recon_loss:0.512107, zinb_loss:0.924047, adversial_loss:1.249946
Pretrain epoch 269, recon_loss:0.514833, zinb_loss:0.925737, adversial_loss:1.250677
Pretrain epoch 270, recon_loss:0.517847, zinb_loss:0.926574, adversial_loss:1.249720
Pretrain epoch 271, recon_loss:0.517943, zinb_loss:0.924881, adversial_loss:1.250278
Pretrain epoch 272, recon_loss:0.514191, zinb_loss:0.922930, adversial_loss:1.250168
Pretrain epoch 273, recon_loss:0.521242, zinb_loss:0.925202, adversial_loss:1.249156
Pretrain epoch 274, recon_loss:0.517442, zinb_loss:0.925336, adversial_loss:1.250511
Pretrain epoch 275, recon_loss:0.518384, zinb_loss:0.925294, adversial_loss:1.249870
Pretrain epoch 276, recon_loss:0.521161, zinb_loss:0.927005, adversial_loss:1.249135
Pretrain epoch 277, recon_loss:0.514256, zinb_loss:0.923433, adversial_loss:1.249893
Pretrain epoch 278, recon_loss:0.515535, zinb_loss:0.923530, adversial_loss:1.249849
Pretrain epoch 279, recon_loss:0.512503, zinb_loss:0.923560, adversial_loss:1.249354
Pretrain epoch 280, recon_loss:0.515305, zinb_loss:0.922990, adversial_loss:1.249665
Pretrain epoch 281, recon_loss:0.509805, zinb_loss:0.922676, adversial_loss:1.249541
Pretrain epoch 282, recon_loss:0.510216, zinb_loss:0.922475, adversial_loss:1.249405
Pretrain epoch 283, recon_loss:0.507578, zinb_loss:0.922051, adversial_loss:1.249196
Pretrain epoch 284, recon_loss:0.506245, zinb_loss:0.922554, adversial_loss:1.249519
Pretrain epoch 285, recon_loss:0.504669, zinb_loss:0.922000, adversial_loss:1.249220
Pretrain epoch 286, recon_loss:0.502729, zinb_loss:0.921837, adversial_loss:1.248951
Pretrain epoch 287, recon_loss:0.501774, zinb_loss:0.921657, adversial_loss:1.249160
Pretrain epoch 288, recon_loss:0.501260, zinb_loss:0.921655, adversial_loss:1.248994
Pretrain epoch 289, recon_loss:0.499360, zinb_loss:0.921550, adversial_loss:1.249006
Pretrain epoch 290, recon_loss:0.498534, zinb_loss:0.921686, adversial_loss:1.248650
Pretrain epoch 291, recon_loss:0.497990, zinb_loss:0.921559, adversial_loss:1.248952
Pretrain epoch 292, recon_loss:0.498962, zinb_loss:0.921546, adversial_loss:1.248601
Pretrain epoch 293, recon_loss:0.500745, zinb_loss:0.921814, adversial_loss:1.248916
Pretrain epoch 294, recon_loss:0.501947, zinb_loss:0.922106, adversial_loss:1.248188
Pretrain epoch 295, recon_loss:0.498946, zinb_loss:0.921893, adversial_loss:1.248592
Pretrain epoch 296, recon_loss:0.496131, zinb_loss:0.921547, adversial_loss:1.248443
Pretrain epoch 297, recon_loss:0.496319, zinb_loss:0.921698, adversial_loss:1.248167
Pretrain epoch 298, recon_loss:0.496336, zinb_loss:0.921654, adversial_loss:1.248361
Pretrain epoch 299, recon_loss:0.496867, zinb_loss:0.921532, adversial_loss:1.248230
Pretrain epoch 300, recon_loss:0.495937, zinb_loss:0.921501, adversial_loss:1.248012
Pretrain epoch 301, recon_loss:0.498234, zinb_loss:0.922171, adversial_loss:1.248614
Pretrain epoch 302, recon_loss:0.503863, zinb_loss:0.923424, adversial_loss:1.247572
Pretrain epoch 303, recon_loss:0.495321, zinb_loss:0.922800, adversial_loss:1.248073
Pretrain epoch 304, recon_loss:0.498044, zinb_loss:0.921611, adversial_loss:1.248363
Pretrain epoch 305, recon_loss:0.494356, zinb_loss:0.921310, adversial_loss:1.247588
Pretrain epoch 306, recon_loss:0.490891, zinb_loss:0.920596, adversial_loss:1.247740
Pretrain epoch 307, recon_loss:0.491591, zinb_loss:0.921038, adversial_loss:1.247965
Pretrain epoch 308, recon_loss:0.490003, zinb_loss:0.921139, adversial_loss:1.247654
Pretrain epoch 309, recon_loss:0.489223, zinb_loss:0.920694, adversial_loss:1.247563
Pretrain epoch 310, recon_loss:0.487935, zinb_loss:0.920680, adversial_loss:1.247849
Pretrain epoch 311, recon_loss:0.486295, zinb_loss:0.920362, adversial_loss:1.247628
Pretrain epoch 312, recon_loss:0.484374, zinb_loss:0.920141, adversial_loss:1.247453
Pretrain epoch 313, recon_loss:0.483633, zinb_loss:0.920165, adversial_loss:1.247648
Pretrain epoch 314, recon_loss:0.482767, zinb_loss:0.920006, adversial_loss:1.247262
Pretrain epoch 315, recon_loss:0.482459, zinb_loss:0.920056, adversial_loss:1.247328
Pretrain epoch 316, recon_loss:0.484225, zinb_loss:0.920175, adversial_loss:1.247292
Pretrain epoch 317, recon_loss:0.487041, zinb_loss:0.920193, adversial_loss:1.247450
Pretrain epoch 318, recon_loss:0.488344, zinb_loss:0.920604, adversial_loss:1.247184
Pretrain epoch 319, recon_loss:0.482620, zinb_loss:0.920529, adversial_loss:1.246988
Pretrain epoch 320, recon_loss:0.479926, zinb_loss:0.920262, adversial_loss:1.247065
Pretrain epoch 321, recon_loss:0.483461, zinb_loss:0.920427, adversial_loss:1.246775
Pretrain epoch 322, recon_loss:0.482188, zinb_loss:0.920349, adversial_loss:1.247039
Pretrain epoch 323, recon_loss:0.480952, zinb_loss:0.920759, adversial_loss:1.246691
Pretrain epoch 324, recon_loss:0.479341, zinb_loss:0.921360, adversial_loss:1.247012
Pretrain epoch 325, recon_loss:0.484663, zinb_loss:0.921780, adversial_loss:1.246749
Pretrain epoch 326, recon_loss:0.482796, zinb_loss:0.921111, adversial_loss:1.246708
Pretrain epoch 327, recon_loss:0.479651, zinb_loss:0.919901, adversial_loss:1.246666
Pretrain epoch 328, recon_loss:0.475342, zinb_loss:0.919579, adversial_loss:1.246480
Pretrain epoch 329, recon_loss:0.477140, zinb_loss:0.920015, adversial_loss:1.246821
Pretrain epoch 330, recon_loss:0.477093, zinb_loss:0.920330, adversial_loss:1.246401
Pretrain epoch 331, recon_loss:0.476431, zinb_loss:0.920010, adversial_loss:1.246523
Pretrain epoch 332, recon_loss:0.473098, zinb_loss:0.919427, adversial_loss:1.246268
Pretrain epoch 333, recon_loss:0.471720, zinb_loss:0.919005, adversial_loss:1.246389
Pretrain epoch 334, recon_loss:0.470440, zinb_loss:0.919068, adversial_loss:1.246033
Pretrain epoch 335, recon_loss:0.468932, zinb_loss:0.918952, adversial_loss:1.246140
Pretrain epoch 336, recon_loss:0.468013, zinb_loss:0.918944, adversial_loss:1.246304
Pretrain epoch 337, recon_loss:0.471349, zinb_loss:0.919161, adversial_loss:1.245710
Pretrain epoch 338, recon_loss:0.470129, zinb_loss:0.919169, adversial_loss:1.246209
Pretrain epoch 339, recon_loss:0.473457, zinb_loss:0.919853, adversial_loss:1.245862
Pretrain epoch 340, recon_loss:0.475349, zinb_loss:0.920651, adversial_loss:1.246094
Pretrain epoch 341, recon_loss:0.474336, zinb_loss:0.920240, adversial_loss:1.245935
Pretrain epoch 342, recon_loss:0.468043, zinb_loss:0.918994, adversial_loss:1.245606
Pretrain epoch 343, recon_loss:0.468887, zinb_loss:0.918801, adversial_loss:1.245953
Pretrain epoch 344, recon_loss:0.467698, zinb_loss:0.918980, adversial_loss:1.245598
Pretrain epoch 345, recon_loss:0.464630, zinb_loss:0.918694, adversial_loss:1.245645
Pretrain epoch 346, recon_loss:0.466442, zinb_loss:0.918619, adversial_loss:1.245765
Pretrain epoch 347, recon_loss:0.462846, zinb_loss:0.918705, adversial_loss:1.245506
Pretrain epoch 348, recon_loss:0.463364, zinb_loss:0.918919, adversial_loss:1.245496
Pretrain epoch 349, recon_loss:0.466677, zinb_loss:0.919709, adversial_loss:1.245865
Pretrain epoch 350, recon_loss:0.463831, zinb_loss:0.919662, adversial_loss:1.245411
Pretrain epoch 351, recon_loss:0.468532, zinb_loss:0.919222, adversial_loss:1.245589
Pretrain epoch 352, recon_loss:0.463319, zinb_loss:0.919005, adversial_loss:1.245463
Pretrain epoch 353, recon_loss:0.464628, zinb_loss:0.918722, adversial_loss:1.245402
Pretrain epoch 354, recon_loss:0.466382, zinb_loss:0.918665, adversial_loss:1.245361
Pretrain epoch 355, recon_loss:0.463741, zinb_loss:0.918424, adversial_loss:1.245294
Pretrain epoch 356, recon_loss:0.456510, zinb_loss:0.917770, adversial_loss:1.245330
Pretrain epoch 357, recon_loss:0.454428, zinb_loss:0.917779, adversial_loss:1.244992
Pretrain epoch 358, recon_loss:0.454976, zinb_loss:0.917737, adversial_loss:1.245449
Pretrain epoch 359, recon_loss:0.454311, zinb_loss:0.917862, adversial_loss:1.244968
Pretrain epoch 360, recon_loss:0.452632, zinb_loss:0.917909, adversial_loss:1.245039
Pretrain epoch 361, recon_loss:0.453067, zinb_loss:0.917687, adversial_loss:1.244955
Pretrain epoch 362, recon_loss:0.456525, zinb_loss:0.917979, adversial_loss:1.245360
Pretrain epoch 363, recon_loss:0.462370, zinb_loss:0.918319, adversial_loss:1.244389
Pretrain epoch 364, recon_loss:0.454625, zinb_loss:0.917606, adversial_loss:1.245090
Pretrain epoch 365, recon_loss:0.446788, zinb_loss:0.917397, adversial_loss:1.244783
Pretrain epoch 366, recon_loss:0.449447, zinb_loss:0.917623, adversial_loss:1.244616
Pretrain epoch 367, recon_loss:0.449955, zinb_loss:0.917494, adversial_loss:1.244825
Pretrain epoch 368, recon_loss:0.448117, zinb_loss:0.917450, adversial_loss:1.244515
Pretrain epoch 369, recon_loss:0.449342, zinb_loss:0.917591, adversial_loss:1.244497
Pretrain epoch 370, recon_loss:0.448283, zinb_loss:0.917546, adversial_loss:1.244743
Pretrain epoch 371, recon_loss:0.445914, zinb_loss:0.917302, adversial_loss:1.244316
Pretrain epoch 372, recon_loss:0.443342, zinb_loss:0.917304, adversial_loss:1.244421
Pretrain epoch 373, recon_loss:0.443727, zinb_loss:0.917965, adversial_loss:1.244472
Pretrain epoch 374, recon_loss:0.450613, zinb_loss:0.920401, adversial_loss:1.244195
Pretrain epoch 375, recon_loss:0.458434, zinb_loss:0.923473, adversial_loss:1.244600
Pretrain epoch 376, recon_loss:0.466628, zinb_loss:0.923974, adversial_loss:1.244055
Pretrain epoch 377, recon_loss:0.448002, zinb_loss:0.918008, adversial_loss:1.244360
Pretrain epoch 378, recon_loss:0.453877, zinb_loss:0.919228, adversial_loss:1.244228
Pretrain epoch 379, recon_loss:0.454729, zinb_loss:0.920082, adversial_loss:1.244211
Pretrain epoch 380, recon_loss:0.445237, zinb_loss:0.917129, adversial_loss:1.244172
Pretrain epoch 381, recon_loss:0.450492, zinb_loss:0.917859, adversial_loss:1.243852
Pretrain epoch 382, recon_loss:0.442563, zinb_loss:0.917322, adversial_loss:1.244273
Pretrain epoch 383, recon_loss:0.441668, zinb_loss:0.916980, adversial_loss:1.244167
Pretrain epoch 384, recon_loss:0.439400, zinb_loss:0.916999, adversial_loss:1.243751
Pretrain epoch 385, recon_loss:0.436498, zinb_loss:0.916621, adversial_loss:1.243968
Pretrain epoch 386, recon_loss:0.434738, zinb_loss:0.916643, adversial_loss:1.243971
Pretrain epoch 387, recon_loss:0.433600, zinb_loss:0.916510, adversial_loss:1.243769
Pretrain epoch 388, recon_loss:0.432909, zinb_loss:0.916476, adversial_loss:1.243826
Pretrain epoch 389, recon_loss:0.433671, zinb_loss:0.916366, adversial_loss:1.243977
Pretrain epoch 390, recon_loss:0.440959, zinb_loss:0.916790, adversial_loss:1.243476
Pretrain epoch 391, recon_loss:0.441706, zinb_loss:0.916533, adversial_loss:1.244172
Pretrain epoch 392, recon_loss:0.437726, zinb_loss:0.916482, adversial_loss:1.243456
Pretrain epoch 393, recon_loss:0.427984, zinb_loss:0.916146, adversial_loss:1.243696
Pretrain epoch 394, recon_loss:0.428263, zinb_loss:0.916296, adversial_loss:1.243844
Pretrain epoch 395, recon_loss:0.431406, zinb_loss:0.916571, adversial_loss:1.243382
Pretrain epoch 396, recon_loss:0.424839, zinb_loss:0.916067, adversial_loss:1.243597
Pretrain epoch 397, recon_loss:0.426324, zinb_loss:0.916186, adversial_loss:1.243752
Pretrain epoch 398, recon_loss:0.432253, zinb_loss:0.916296, adversial_loss:1.243196
Pretrain epoch 399, recon_loss:0.433320, zinb_loss:0.916364, adversial_loss:1.243449
Pretrain epoch 400, recon_loss:0.431458, zinb_loss:0.916175, adversial_loss:1.243666
[19]:
y_pred, final_latent = model.fit(y=y, n_clusters=5, file='10X1kpbmc')
Clustering stage
Initializing cluster centers with kmeans.
Initializing k-means: AMI= 0.8205, NMI= 0.8218, ARI= 0.8344, ACC= 0.9201
Training epoch 1, recon_loss:0.430522, zinb_loss:0.916023, cluster_loss:0.154342
Clustering 1: AMI= 0.8205, NMI= 0.8218, ARI= 0.8344, ACC= 0.9201
0.0
Training epoch 2, recon_loss:0.846306, zinb_loss:1.077583, cluster_loss:0.171737
Clustering 2: AMI= 0.8006, NMI= 0.8021, ARI= 0.8141, ACC= 0.9088
0.014025245441795231
Training epoch 3, recon_loss:0.778181, zinb_loss:1.139828, cluster_loss:0.164231
Clustering 3: AMI= 0.8304, NMI= 0.8316, ARI= 0.8507, ACC= 0.9341
0.03225806451612903
Training epoch 4, recon_loss:0.784476, zinb_loss:1.364775, cluster_loss:0.169950
Clustering 4: AMI= 0.8110, NMI= 0.8124, ARI= 0.8221, ACC= 0.9159
0.03927068723702665
Training epoch 5, recon_loss:0.830803, zinb_loss:1.707875, cluster_loss:0.176492
Clustering 5: AMI= 0.8033, NMI= 0.8047, ARI= 0.7921, ACC= 0.8780
0.06591865357643759
Training epoch 6, recon_loss:0.802670, zinb_loss:1.205563, cluster_loss:0.177979
Clustering 6: AMI= 0.8130, NMI= 0.8144, ARI= 0.8361, ACC= 0.9229
0.08134642356241234
Training epoch 7, recon_loss:0.773048, zinb_loss:1.173137, cluster_loss:0.170710
Clustering 7: AMI= 0.8174, NMI= 0.8187, ARI= 0.8373, ACC= 0.9243
0.005610098176718092
Training epoch 8, recon_loss:0.715895, zinb_loss:1.001829, cluster_loss:0.163364
Clustering 8: AMI= 0.8115, NMI= 0.8128, ARI= 0.8306, ACC= 0.9186
0.008415147265077139
Training epoch 9, recon_loss:0.707180, zinb_loss:0.976355, cluster_loss:0.150441
Clustering 9: AMI= 0.8095, NMI= 0.8109, ARI= 0.8083, ACC= 0.9018
0.037868162692847124
Training epoch 10, recon_loss:0.661050, zinb_loss:0.958861, cluster_loss:0.151975
Clustering 10: AMI= 0.8213, NMI= 0.8226, ARI= 0.8308, ACC= 0.9186
0.019635343618513323
Training epoch 11, recon_loss:0.652488, zinb_loss:0.952982, cluster_loss:0.149777
Clustering 11: AMI= 0.8103, NMI= 0.8116, ARI= 0.8189, ACC= 0.9102
0.011220196353436185
Training epoch 12, recon_loss:0.642588, zinb_loss:0.953146, cluster_loss:0.151356
Clustering 12: AMI= 0.8215, NMI= 0.8228, ARI= 0.8335, ACC= 0.9201
0.009817671809256662
Training epoch 13, recon_loss:0.638285, zinb_loss:0.949473, cluster_loss:0.149443
Clustering 13: AMI= 0.8164, NMI= 0.8177, ARI= 0.8254, ACC= 0.9159
0.0070126227208976155
Training epoch 14, recon_loss:0.641145, zinb_loss:0.953004, cluster_loss:0.147145
Clustering 14: AMI= 0.8178, NMI= 0.8192, ARI= 0.8276, ACC= 0.9159
0.002805049088359046
Training epoch 15, recon_loss:0.637767, zinb_loss:0.952127, cluster_loss:0.148271
Clustering 15: AMI= 0.8203, NMI= 0.8216, ARI= 0.8315, ACC= 0.9186
0.002805049088359046
Training epoch 16, recon_loss:0.625828, zinb_loss:0.957454, cluster_loss:0.145889
Clustering 16: AMI= 0.8140, NMI= 0.8154, ARI= 0.8209, ACC= 0.9116
0.009817671809256662
Training epoch 17, recon_loss:0.611839, zinb_loss:0.962478, cluster_loss:0.148376
Clustering 17: AMI= 0.8260, NMI= 0.8273, ARI= 0.8380, ACC= 0.9229
0.015427769985974754
Training epoch 18, recon_loss:0.621567, zinb_loss:0.977118, cluster_loss:0.145853
Clustering 18: AMI= 0.8140, NMI= 0.8154, ARI= 0.8209, ACC= 0.9116
0.015427769985974754
Training epoch 19, recon_loss:0.607203, zinb_loss:0.988093, cluster_loss:0.149063
Clustering 19: AMI= 0.8275, NMI= 0.8288, ARI= 0.8401, ACC= 0.9257
0.021037868162692847
Training epoch 20, recon_loss:0.615731, zinb_loss:0.995491, cluster_loss:0.145251
Clustering 20: AMI= 0.8140, NMI= 0.8154, ARI= 0.8209, ACC= 0.9116
0.021037868162692847
Training epoch 21, recon_loss:0.609800, zinb_loss:0.997153, cluster_loss:0.147204
Clustering 21: AMI= 0.8261, NMI= 0.8274, ARI= 0.8380, ACC= 0.9243
0.019635343618513323
Training epoch 22, recon_loss:0.621488, zinb_loss:0.992326, cluster_loss:0.146616
Clustering 22: AMI= 0.8182, NMI= 0.8195, ARI= 0.8240, ACC= 0.9130
0.0182328190743338
Training epoch 23, recon_loss:0.611237, zinb_loss:0.988998, cluster_loss:0.145771
Clustering 23: AMI= 0.8203, NMI= 0.8216, ARI= 0.8315, ACC= 0.9186
0.008415147265077139
Training epoch 24, recon_loss:0.600035, zinb_loss:0.983440, cluster_loss:0.144918
Clustering 24: AMI= 0.8253, NMI= 0.8265, ARI= 0.8355, ACC= 0.9215
0.002805049088359046
Training epoch 25, recon_loss:0.596902, zinb_loss:0.978649, cluster_loss:0.143721
Clustering 25: AMI= 0.8182, NMI= 0.8195, ARI= 0.8240, ACC= 0.9130
0.011220196353436185
Training epoch 26, recon_loss:0.606475, zinb_loss:0.980079, cluster_loss:0.145489
Clustering 26: AMI= 0.8304, NMI= 0.8316, ARI= 0.8440, ACC= 0.9285
0.025245441795231416
Training epoch 27, recon_loss:0.620056, zinb_loss:0.975876, cluster_loss:0.146040
Clustering 27: AMI= 0.8160, NMI= 0.8173, ARI= 0.8204, ACC= 0.9102
0.028050490883590462
Training epoch 28, recon_loss:0.627696, zinb_loss:0.977644, cluster_loss:0.149002
Clustering 28: AMI= 0.8264, NMI= 0.8277, ARI= 0.8437, ACC= 0.9285
0.0364656381486676
Training epoch 29, recon_loss:0.607567, zinb_loss:0.963646, cluster_loss:0.145189
Clustering 29: AMI= 0.8171, NMI= 0.8184, ARI= 0.8222, ACC= 0.9116
0.03506311360448808
Training epoch 30, recon_loss:0.596643, zinb_loss:0.960702, cluster_loss:0.144895
Clustering 30: AMI= 0.8280, NMI= 0.8292, ARI= 0.8395, ACC= 0.9257
0.019635343618513323
Training epoch 31, recon_loss:0.586386, zinb_loss:0.957384, cluster_loss:0.143485
Clustering 31: AMI= 0.8190, NMI= 0.8204, ARI= 0.8296, ACC= 0.9173
0.011220196353436185
Training epoch 32, recon_loss:0.587456, zinb_loss:0.956871, cluster_loss:0.142214
Clustering 32: AMI= 0.8189, NMI= 0.8202, ARI= 0.8294, ACC= 0.9186
0.004207573632538569
Training epoch 33, recon_loss:0.586300, zinb_loss:0.957902, cluster_loss:0.142721
Clustering 33: AMI= 0.8203, NMI= 0.8216, ARI= 0.8315, ACC= 0.9186
0.002805049088359046
Training epoch 34, recon_loss:0.597041, zinb_loss:0.956997, cluster_loss:0.141408
Clustering 34: AMI= 0.8152, NMI= 0.8166, ARI= 0.8235, ACC= 0.9144
0.004207573632538569
Training epoch 35, recon_loss:0.582191, zinb_loss:0.959904, cluster_loss:0.142487
Clustering 35: AMI= 0.8228, NMI= 0.8241, ARI= 0.8355, ACC= 0.9215
0.0070126227208976155
Training epoch 36, recon_loss:0.579865, zinb_loss:0.960015, cluster_loss:0.141827
Clustering 36: AMI= 0.8166, NMI= 0.8180, ARI= 0.8217, ACC= 0.9130
0.011220196353436185
Training epoch 37, recon_loss:0.579633, zinb_loss:0.963965, cluster_loss:0.142911
Clustering 37: AMI= 0.8275, NMI= 0.8288, ARI= 0.8401, ACC= 0.9257
0.016830294530154277
Training epoch 38, recon_loss:0.591656, zinb_loss:0.964938, cluster_loss:0.143914
Clustering 38: AMI= 0.8103, NMI= 0.8117, ARI= 0.8149, ACC= 0.9088
0.021037868162692847
Training epoch 39, recon_loss:0.598165, zinb_loss:0.971153, cluster_loss:0.144402
Clustering 39: AMI= 0.8260, NMI= 0.8273, ARI= 0.8394, ACC= 0.9257
0.02244039270687237
Training epoch 40, recon_loss:0.586254, zinb_loss:0.966595, cluster_loss:0.142439
Clustering 40: AMI= 0.8155, NMI= 0.8169, ARI= 0.8199, ACC= 0.9116
0.019635343618513323
Training epoch 41, recon_loss:0.576397, zinb_loss:0.964681, cluster_loss:0.141448
Clustering 41: AMI= 0.8202, NMI= 0.8215, ARI= 0.8314, ACC= 0.9201
0.011220196353436185
Training epoch 42, recon_loss:0.580564, zinb_loss:0.965172, cluster_loss:0.142515
Clustering 42: AMI= 0.8164, NMI= 0.8177, ARI= 0.8254, ACC= 0.9159
0.004207573632538569
Training epoch 43, recon_loss:0.581782, zinb_loss:0.966008, cluster_loss:0.141712
Clustering 43: AMI= 0.8178, NMI= 0.8192, ARI= 0.8276, ACC= 0.9159
0.002805049088359046
Training epoch 44, recon_loss:0.578958, zinb_loss:0.961043, cluster_loss:0.142084
Clustering 44: AMI= 0.8202, NMI= 0.8215, ARI= 0.8314, ACC= 0.9201
0.0070126227208976155
Training epoch 45, recon_loss:0.578888, zinb_loss:0.958669, cluster_loss:0.141906
Clustering 45: AMI= 0.8182, NMI= 0.8195, ARI= 0.8240, ACC= 0.9130
0.012622720897615708
Training epoch 46, recon_loss:0.572623, zinb_loss:0.953206, cluster_loss:0.141412
Clustering 46: AMI= 0.8215, NMI= 0.8228, ARI= 0.8334, ACC= 0.9215
0.014025245441795231
Training epoch 47, recon_loss:0.573183, zinb_loss:0.951851, cluster_loss:0.141826
Clustering 47: AMI= 0.8171, NMI= 0.8184, ARI= 0.8222, ACC= 0.9116
0.015427769985974754
Training epoch 48, recon_loss:0.582879, zinb_loss:0.955231, cluster_loss:0.141006
Clustering 48: AMI= 0.8252, NMI= 0.8265, ARI= 0.8354, ACC= 0.9229
0.016830294530154277
Training epoch 49, recon_loss:0.577148, zinb_loss:0.950773, cluster_loss:0.141801
Clustering 49: AMI= 0.8182, NMI= 0.8195, ARI= 0.8240, ACC= 0.9130
0.015427769985974754
Training epoch 50, recon_loss:0.582462, zinb_loss:0.950802, cluster_loss:0.140044
Clustering 50: AMI= 0.8189, NMI= 0.8202, ARI= 0.8294, ACC= 0.9186
0.011220196353436185
Training epoch 51, recon_loss:0.578021, zinb_loss:0.947385, cluster_loss:0.141623
Clustering 51: AMI= 0.8176, NMI= 0.8190, ARI= 0.8274, ACC= 0.9173
0.001402524544179523
Training epoch 52, recon_loss:0.593130, zinb_loss:0.945707, cluster_loss:0.141130
Clustering 52: AMI= 0.8155, NMI= 0.8169, ARI= 0.8199, ACC= 0.9116
0.008415147265077139
Training epoch 53, recon_loss:0.584805, zinb_loss:0.945096, cluster_loss:0.142527
Clustering 53: AMI= 0.8202, NMI= 0.8215, ARI= 0.8314, ACC= 0.9201
0.011220196353436185
Training epoch 54, recon_loss:0.585467, zinb_loss:0.943358, cluster_loss:0.140558
Clustering 54: AMI= 0.8144, NMI= 0.8158, ARI= 0.8180, ACC= 0.9102
0.012622720897615708
Training epoch 55, recon_loss:0.568052, zinb_loss:0.941190, cluster_loss:0.141217
Clustering 55: AMI= 0.8202, NMI= 0.8215, ARI= 0.8314, ACC= 0.9201
0.012622720897615708
Training epoch 56, recon_loss:0.577989, zinb_loss:0.941180, cluster_loss:0.139027
Clustering 56: AMI= 0.8117, NMI= 0.8131, ARI= 0.8154, ACC= 0.9088
0.014025245441795231
Training epoch 57, recon_loss:0.572676, zinb_loss:0.940440, cluster_loss:0.141063
Clustering 57: AMI= 0.8242, NMI= 0.8255, ARI= 0.8375, ACC= 0.9243
0.0182328190743338
Training epoch 58, recon_loss:0.580458, zinb_loss:0.941732, cluster_loss:0.139396
Clustering 58: AMI= 0.8117, NMI= 0.8131, ARI= 0.8154, ACC= 0.9088
0.0182328190743338
Training epoch 59, recon_loss:0.564884, zinb_loss:0.940472, cluster_loss:0.141210
Clustering 59: AMI= 0.8215, NMI= 0.8228, ARI= 0.8334, ACC= 0.9215
0.015427769985974754
Training epoch 60, recon_loss:0.566060, zinb_loss:0.942304, cluster_loss:0.139360
Clustering 60: AMI= 0.8144, NMI= 0.8158, ARI= 0.8180, ACC= 0.9102
0.014025245441795231
Training epoch 61, recon_loss:0.561706, zinb_loss:0.942914, cluster_loss:0.141014
Clustering 61: AMI= 0.8202, NMI= 0.8215, ARI= 0.8314, ACC= 0.9201
0.012622720897615708
Training epoch 62, recon_loss:0.577613, zinb_loss:0.950187, cluster_loss:0.139490
Clustering 62: AMI= 0.8155, NMI= 0.8169, ARI= 0.8199, ACC= 0.9116
0.011220196353436185
Training epoch 63, recon_loss:0.569885, zinb_loss:0.951495, cluster_loss:0.140008
Clustering 63: AMI= 0.8176, NMI= 0.8190, ARI= 0.8274, ACC= 0.9173
0.008415147265077139
Training epoch 64, recon_loss:0.569891, zinb_loss:0.955939, cluster_loss:0.139665
Clustering 64: AMI= 0.8176, NMI= 0.8190, ARI= 0.8274, ACC= 0.9173
0.0
delta_label 0.0 < tol 0.001
Reach tolerance threshold. Stopping training.
Final Result : AMI= 0.8176, NMI= 0.8190, ARI= 0.8274, ACC= 0.9173
[ ]:
import numpy as np
y_pred_ = best_map(y, y_pred).astype(int)
np.savetxt("../results/10X1kpbmc_pred.csv", y_pred, delimiter=",")
np.savetxt("../results/10X1kpbmc_embedding.csv", final_latent.cpu().detach().numpy(), delimiter=",")
[41]:
library(umap)
library(ggplot2)
library(scattermore)
[42]:
latent <- read.csv(file = "../results/10X1kpbmc_embedding.csv", sep = ",", header = FALSE)
y_pred <- read.csv(file = "../results/10X1kpbmc_pred.csv", sep = ",", header = FALSE)
[43]:
z.umap<-umap(latent)
y_pred$V1<-as.factor(y_pred$V1)
[44]:
bm_table_num <- data.frame(UMAP1=z.umap$layout[,1], UMAP2=z.umap$layout[,2],Cluster = y_pred$V1)
bm_plot1 <- ggplot(bm_table_num, aes(x = UMAP1, y = UMAP2, color = Cluster)) +
geom_point(size = 0.1) +
ggtitle('scMAGCA') +
theme_bw(base_line_size = 1,base_rect_size = 1)+
scale_color_manual(values = c("1"="#e0bc58", "2"="#64abc0", "3"="#dc8e97", "4"="#74a893", "5"="#ac9141"))+
labs(x = "", y = "", color = "Label")+ # Set x and y axis labels
theme(
plot.title = element_text(size = 20, hjust = 0.5, face = 'bold'), # Set title to center
panel.border = element_blank(),
axis.ticks.length=unit(0, "lines"), # Remove axis ticks but keep axis labels
axis.text = element_blank(),
panel.grid.minor = element_blank(),
panel.grid.major = element_line(color = NA),
axis.title.x = element_text(size = 15, hjust = 0.5, vjust = 0.5, color = "black"),
axis.title.y = element_text(size = 15, hjust = 0.5, vjust = 0.5, color = "black"),
legend.title = element_blank(),
legend.text = element_text(size = 15),
legend.position = "none")+
guides(color = guide_legend(override.aes = list(size = 5), nrow=2))
[45]:
bm_plot1
