nanmean error addressed added h5 reformmating script

chrombpnet/helpers/postprocessing/reformat_chrombpnet_h5.py

@@ -0,0 +1,60 @@
+from tensorflow.keras.models import load_model
+from tensorflow.keras.utils import get_custom_objects
+import argparse
+import tensorflow as tf
+import numpy as np
+from tensorflow.keras.models import Model
+from tensorflow.keras.layers import Input, Cropping1D, add, Conv1D, GlobalAvgPool1D, Dense, Add, Concatenate, Lambda, Flatten
+import time
+import os
+import argparse
+
+def parse_args():
+        parser = argparse.ArgumentParser(description="Reformat chrombpnet h5 file")
+        parser.add_argument("-cnb", "--chrombpnet_nb", type=str, required=True, help="Path to chrombpnet no bias model")
+        parser.add_argument("-bm", "--bias_model_scaled", type=str, required=True, help="Path to scaled bias model")        
+        parser.add_argument("-o", "--output_dir", type=str, required=True, help="Path to output dir")        
+        args = parser.parse_args()
+        return args
+
+
+args = parse_args()
+
+args_chrombpnet_nb=args.chrombpnet_nb
+args_bias=args.bias_model_scaled
+args_output_dir=args.output_dir
+
+def chrombpnet_model(bias_model, bpnet_model_wo_bias):
+	inp = Input(shape=(2114, 4),name='sequence')    
+	bias_output=bias_model(inp)
+	bpnet_model_wo_bias_new=Model(inputs=bpnet_model_wo_bias.inputs,outputs=bpnet_model_wo_bias.outputs, name="model_wo_bias")
+	output_wo_bias=bpnet_model_wo_bias_new(inp)
+
+	profile_out = Add(name="logits_profile_predictions")([output_wo_bias[0],bias_output[0]])
+	concat_counts = Concatenate(axis=-1)([output_wo_bias[1], bias_output[1]])
+	count_out = Lambda(lambda x: tf.math.reduce_logsumexp(x, axis=-1, keepdims=True),
+						name="logcount_predictions")(concat_counts)
+	model=Model(inputs=[inp],outputs=[profile_out, count_out])
+	return model
+
+def main(args_chrombpnet_nb, args_bias, args_output_dir):
+
+	custom_objects={"tf":tf}  
+	get_custom_objects().update(custom_objects)
+
+	chrombpnet_nb=load_model(args_chrombpnet_nb,compile=False)
+	bias_model=load_model(args_bias,compile=False)
+
+	newp = args_output_dir+"/chrombpnet_recompiled.h5"
+	new_chrom = chrombpnet_model(bias_model, chrombpnet_nb)
+	new_chrom.save(newp)
+	newp = args_output_dir+"/chrombpnet_recompiled"
+	new_chrom.save(newp)
+
+if __name__ == '__main__':
+
+		main(args_chrombpnet_nb, args_bias, args_output_dir)
+
+
+
+

chrombpnet/training/predict.py

@@ -11,7 +11,7 @@
 import chrombpnet.training.data_generators.initializers as initializers
 from tensorflow.keras.utils import get_custom_objects
 from tensorflow.keras.models import load_model
-from scipy import nanmean, nanstd
+#from scipy import nanmean, nanstd
 
 def write_predictions_h5py(output_prefix, profile, logcts, coords):
     # open h5 file for writing predictions