#!/bin/bash
echo "enter the file name with it's extintion"
read file
if [ ! -f "$file" ]
then
echo "file dose not exist"
exit 1
fi


if [  ! $( echo $file | grep '\.csv$' ) ]
then
echo "wrong file format (the extintion is not .csv)"
exit 1
fi

rows=$(cat $file | wc -l)
rows=$(( rows - 1 ))

#calculate the number of coloms by calcuate the first line comas
coloms=$(sed -n '1p' $file | awk -F, '{print NF-1}')
coloms=$(( coloms + 1 ))

#send the first line string into another file
touch titels
touch nums
cat $file | head -1 > titels

#send the numbers into file
cat $file | tail -$rows > nums
mv nums $file


while IFS= read line
do

#erorr messege if the file contain somtheng other than numbers or commas
if [ $(echo "$line" | grep  [A-Za-z?@#+'&'*$%^/!\-]) ]
then
echo file format is wrong fix it please
exit 1
fi

lineCo=$(echo $line | awk -F, '{print NF-1}')
lineCo=$(( lineCo + 1 ))


if [ $lineCo -ne $coloms ]
then
echo "Error: number of cells are not equal in all rows."
exit 1
fi

done < "$file"


for (( i=1; i<=$coloms; i++ ))
do

touch $i

done


#seperate every colom into a file
while IFS= read line
do

for (( i=1; i<=$coloms; i++ ))
do



num=$(echo $line | cut -d ',' -f$i)

echo $num >> $i

done

done < "$file"

#infinte loop to repeat the orders
while [ true ]
do
echo "please chose the character of the action below"
echo " D : for Dimension"
echo " C : for computing statistics"
echo " S : for subsitution"
echo " X : for exit"
read input

#the dimention option
if [ "$input" = D -o "$input" = d ]
then

echo "the dimention is $rows X $coloms"




#the calculate statistics option max, min, mean, sdev
elif [ "$input" = C -o "$input" = c ]
then

printf "max "
for (( f=1; f<=$coloms; f++ ))
do

#it grep the lines and sort them then take the first line as the greatest number
max=$(cat "$f" | grep ^[-0-9] | sort -gr | head -1)
printf "%f    " $max

done
echo



printf "min "
for (( f=1; f<=$coloms; f++ ))
do
#it grep the lines and sort them in reverse order then take the first line as the minimum number
min=$(cat "$f" | grep ^[-0-9] | sort -g | head -1)
printf "%f    " $min

done
echo


#print the mean after calculate it for every file
printf "mean "
for (( f=1; f<=$coloms; f++ ))
do

mean=0
sum=0
count=$(cat $f | wc -l)


while IFS= read line
do

if [ -z "$line" ]
then
count=$(( count - 1 ))


elif [ $(echo $line | grep '[0-9]') ]
then
sum=$(calc $line + $sum )

fi

done < "$f"

#take the first4 digits after the point
mean=$(echo "scale=4; $sum / $count" | bc)

printf "%f    " $mean


done
echo

# calculte the stander diveation
printf "sdev "
for (( f=1; f<=$coloms; f++ ))
do


mean=0
sum=0
sd=0
sum2=0
# ignore the empty lines
count=$(cat $f | grep '[0-9]' | wc -l)
while IFS= read line
do

sum=$(calc $line + $sum )

done < "$f"

mean=$(calc $sum / $count)




while IFS= read line
do

lm=$(calc $line - $mean)


lm2=$( echo $lm \* $lm | bc )


sum2=$(calc $sum2 + $lm2)


done < "$f"

bf=$(echo "scale=2; $sum2 / $count" | bc)

sd=$(echo "$bf" | awk '{print sqrt($1)}')

printf "%f    " $sd

done

echo





elif [ "$input" = S -o "$input" = s ]
then



for (( i=1; i<=$coloms; i++ ))
do

mean=0
sum=0
count=$(cat $i | wc -l)


while IFS= read line
do

if [ -z "$line" ]
then
count=$(( count - 1 ))


else

sum=$(calc $line + $sum )

fi

done < "$i"

mean=$(echo "scale=2; $sum / $count" | bc)


while IFS= read line
do
#subsitute the mean in the empty cells
if [ -z $line ]
then
echo $mean >> temp
else
echo $line >> temp
fi

done < "$i"

mv temp $i
done


#marege the seperated files to the mother file again
x=1
for (( i=2; i<=$coloms; i++ ))
do
x="$x $i"
done

paste -d ',' $x > $file
cat "$file" >> titels
mv titels "$file"


elif [ "$input" = X -o "$input" = x ]
then
echo thank you

#delete the used files for calculations
x=1
for (( i=2; i<=$coloms; i++ ))
do
x="$x $i"
done
rm  $x

exit 1

else
echo wrong character try again

fi

done