run_fire_components_ens.sh

#!/bin/bash

VARS=(  gfedl1  prl1  prl2  gppl1  gppl2  pr  ts  cld  vp  pop  rdtot   )

ARR=('USEV=(       0     0     0      0      0   1   1    1   1    1      1   )'
	 'USEV=(       0     0     0      0      0   0   0    1   1    1      1   )')

for i in "${ARR[@]}"; do
eval $i

FOLDER=output_globe
MODEL=NHAF

## Generate a unique code number for the model from the variables used.
MODNUM="${USEV[@]}" 			# join USEV array serially (joins with spaces)
MODNUM=$((2#${MODNUM// /}))		# remove spaces so MODNUM becomes a binary number, then convert to decimal
MODEL=${MODEL}_$MODNUM.0		# Append decimal number to model name. Thus each model gets unique name
echo MODEL CODE=$MODEL

done 

##################################################################
##	Update code, folder names etc as per specified variables    #
##################################################################

#N=${#VARS[@]}		# get length of variables array
#VARS_LIST=""		# Create a text list of varnames to be used by C++ nc2asc 
#XID="X_ids = ["		# Create a line of tensorflow code that defines training variables
#for ((i=0; i<$N; ++i)); do
##	echo ${VARS[$i]}
#	if ((${USEV[$i]} == 1)); then
#		VARS_LIST=${VARS_LIST}${VARS[$i]}"\n"
#		V=$(sed "s/_//g" <<< ${VARS[$i]})	# remove underscore from variable name (for folder naming)
#		MODEL=${MODEL}_$V					# append variable to model name
#		XID=${XID}"ID_${VARS[$i]}, " 		# append to the X_ids array for tensorflow code
#	fi
#done
#echo MODEL = $MODEL
#XID=$(sed '$ s/.$//' <<< $XID)				# remove the trailing comma
#XID=${XID}"]" 								# Add closing bracket

#XIDLINE=$(grep -rn "X\_ids \= " tensorflow/nn_const_data_fire_v5_pureNN.py | cut -f1 -d:)	# Get the line that defines X_ids in the tensorflow code file

#echo "Replace Line $XIDLINE:" 
#echo -e "\033[0;31m- " $(grep -r "X\_ids \= " tensorflow/nn_const_data_fire_v5_pureNN.py)
#echo -e "\033[0;32m+ " $XID "\033[0m"

#sed -i "${XIDLINE}s/.*/${XID}/" tensorflow/nn_const_data_fire_v5_pureNN.py	# Replace this line with the newly created X_ids 

#mkdir -p $FOLDER/$MODEL

#echo -e $VARS_LIST > $FOLDER/$MODEL/nn_vars.txt		# Output the list of vars to file (to be used by nc2asc)

#make 

###################################################################
###	Run fire components                                         #
###################################################################

###### Aggregate training data
#####./nc2asc train params_newdata/params_ip_global.r
#####Rscript Rscripts/prepare_train_eval_datasets.R

### Train NN
##cd tensorflow
##. runtf 
##cd ..

### Run trained NN on data
### ## nc2asc eval syntax: ./nc2asc eval <params_file> <model_dir> <weights_file> <vars_file>
##./nc2asc eval params_newdata/params_ip_global.r $FOLDER/$MODEL weights_ba.txt nn_vars.txt
#####mv fire.2003-1-1-2015-12-31.nc $FOLDER/$MODEL

### Plot results
##cd Rscripts
##Rscript plot_aggregate_maps_timeseries.R model_dir=$MODEL output_dir=$FOLDER

##cd ..