Feed RSS
Feed RSS
Debian GNU Linux
Play Ogg
Eliminate DRM!
hosted by tuxfamily.org
Articles Hierarchy

    Floating point in Bash

A very well known problem of Bash is the absence of floating point number capability. In general shells are not designed to perform calculations and the only that as far as I know can do these operations is the Korn shell 93 (Ksh93). However solutions are widely available even if they are not suitable for long and complex calculations. Short operations are fine anyway. The main problem is that in Bash everything is a string, even numbers, and is impossible to operate on the string representing a decimal number. So we need to rely on external tools. Designed for this purpouse is bc, but also awk could be used.

Bc is not only an executable, is the interpreter of a real programming language able to understand operations passed as strings and give the result. That means that we can use it also for more complex thing than simple additions or multiplications. A detailed guide to bc is not the purpouse of this article, so I will simply scratch the top and show some code to use the simplest features of this interpreter.

Floating point operations

To perform a simple floating point operation pass a string describing the math to bc

$ echo "1.452 + 9.236" | bc


$ echo "171/98" | bc

One is the correct integer part of the result, but we want decimals! bc by default gives us in the output the same number of decimal digits used in the input. The scale parameter manage the number of decimals we want to see

$ echo "scale=5; 171/98" | bc

Operations can be concatenated

$ echo "1.2+3.4*5.6/7.8" | bc

Last result can be reused using last or a dot .

$ echo "1.2+3.4;last*5.6" | bc


$ echo "1.2+3.4;.*5.6" | bc


We can do comparisons with bc

$ echo "1.5 < 4.5" | bc -l
1   # True
$ echo "1.5 > 4.5" | bc -l
0   # False
$ echo "2.4/4.8 == 0.5" | bc -l
1   # True

Use in scripts

Using it in scripts is very easy. Few pieces of codes using bc:
Define a variable. A variable can be defined as usual

FLOAT_NUM=$(echo "scale=5; 171/98" | bc)

A quick function could be

floatCalc() {
    echo "scale=5; $@" | bc -l


floatCalc() {
    bc -l <<< "scale=5; $@"

True/False comparisons can be used inside conditional statements in scripts

if [ $(echo "$1 > $2" | bc -l) -ne 0 ]; then
    echo "True"
    echo "False"

Numbers converters

bc can be used to convert numbers from one base to another. The two special variables ibase and obase define the base of the input and of the output. For example to convert the number 42 from base 10 to binary

$ echo "ibase=10; obase=2; 42" | bc

More operators

Many operators are available using the standard math library enabled using the -l flag. Refer to the bc documentation for a full list. Here a couple of examples.


$ echo "1.5123^9" | bc -l

Natural logaritm:

$ echo "l(1.5123)" | bc -l

Sin, cosine:

$ echo "s(1.5123)" | bc -l
$ echo "c(1.5123)" | bc -l

Square root:

$ echo "sqrt(2)" | bc -l

Final note

There are several flavours of shell for Unix. None of them support floating point numbers with one exception. The Korn shell 93 supports floating-point numbers, but as far as I know is the only version of Ksh with this feature.

Useful websites:

Bc manual

No Comments have been Posted.

Post Comment
Please Login to Post a Comment.

Rating is available to Members only.

Please login to vote.

No Ratings have been Posted.