ec: An Engineering Calculator
This calculator is noteworthy in that it employs a stack model of computation (Reverse Polish Notation), it supports numbers with SI scale factors and units. and uses a text-only user interface.
Installing
Requires Python version 3.6 or later.
Install with:
pip3 install engineering-calculator
This installs ec into ~/.local/bin, which should be added to your path.
Unusually, there is also a man page. The Python install process no longer
supports man pages, however you can download it from GitHub. Place it
in ~/.local/man/man1.
More information on both ec can be found on ReadTheDocs .
Installing from Source
To get the source code:
$ git clone https://github.com/KenKundert/ec.git
Once cloned, you can get the latest updates using:
$ cd ec
$ git pull
To run the regression tests:
$ tox
To install:
$ pip install .
To run EC:
$ ec
0:
Installing Man Page
If you have installed from source, you can install the manpage with:
cd doc
make publish
Otherwise, you can install the latest version of the manpage on GitHub using:
curl https://raw.githubusercontent.com/KenKundert/ec/master/install-manpage | bash -
Once installed, you can get access the man page using:
man ec
A Brief Tour of Engineering Calculator
To perform operations in EC, you first enter the numbers, then the operators. In particular, as you enter the numbers they are pushed onto the stack. The operators then take numbers from the stack and replace them with the result. The operations are performed immediately and there is no use of parentheses to group calculations. Any intermediate results are stored on the stack until needed.
To add two numbers:
0: 4 5 +
9:
This command first pushes 4 onto the stack, then it pushes 5 on the stack, and finally runs the addition operator, which pulls 4 and 5 off the stack and then pushes the sum, 9, back onto the stack. The prompt displays the value of the x-register, which is generally the final result from the previous command.
You can string together an arbitrarily long calculation on a single line:
0: 4 5 + 6 7 + *
117:
This command demonstrates the power of using a stack for calculations. It first computes the sum and places the results on the stack. That result stays on the stack while the sum of 6 and 7 is computed, and finally it is used, and consumed, in the final multiplication.
Alternately, you can string a calculation over multiple lines (this calculates the value of two parallel 100 ohm resistors):
0: 100
100: 100
100: ||
50:
Effectively, you only need to type enter is when you want to see the result.
Select operators can be entered without preceding them with a space if they follow a number or a name. For example:
0: 4 5* 6 5+ *
220:
Use stack to see the contents of the stack:
0: 1 2 3 4 5 stack
1
2
3
y: 4
x: 5
5: + stack
1
2
y: 3
x: 9
9: + stack
1
y: 2
x: 12
12: + stack
y: 1
x: 14
14: + stack
x: 15
14: -1 stack
y: 15
x: -1
-1:
The stack grows without limit as needed. The bottom two values are the values that are generally involved in operations and they are labeled x and y as an aid to help you understand and predict the basic operation of various commands. For example:
0: 8 2 stack
y: 8
x: 2
2: ytox
64:
The command name ytox is short for ‘raise value of y register to the value in the x register’.
You remove a value from the bottom of the stack with pop:
0: 10 -3 stack
y: 10
x: -3
-3: pop
10: stack
x: 10
To store a value into a variable, type an equal sign followed by a name. To recall it, simply use the name:
0: 100MHz =freq
100MHz: 2pi* =omega
628.32M: 1pF =Cin
1pF: 1 omega/ Cin/
1.5915K:
Display variables using:
628.32M: vars
Cin = 1pF
Rref = 50 Ohms
freq = 100MHz
omega = 628.32M
628.32M:
Rref is a special variable that is set by default to 50 Ohms, but you can change its value. It is used in dBm calculations.
From the above example you can see that EC supports SI scale factors and units. The support for units is relatively conservative. You can enter them and it remembers them, but they do not survive any operation other than a copy. In this way it should never display incorrect or misleading units, however it displays units when it can. For example:
0: 100MHz =freq
100 MHz: 2pi* "rads/s" =omega
628.32 Mrads/s: vars
Rref = 50 Ohms
freq = 100 MHz
omega = 628.32 Mrads/s
628.32 Mrads/s: 2pi /
100M:
Notice that EC captured units on 100MHz and stored them into the memory freq. Also notice that the units of “rads/s” were explicitly specified, and they were also captured. Finally, notice that dividing by 2pi cleared the units.
This simple way of adding units to a number, ex. 100MHz, is somewhat restricted.
You can only add units after a scale factor, but once you’ve given the scale factor the units are optional. In this way, 1m represents 1e-3 rather than one meter. If you want to specify one meter, you would use 1_m. The underscore is a scale factor, like m or k. It represents the unity scale factor.
Units added to the end of a number may consist only of letters and underscores. Digits and special characters like /, ^, *, -, ( or ) are not allowed.
You can only add units to number literals. So 100MHz is okay, but ‘omega 2pi/ Hz’ is not.
You can overcome this limitation by entering a quoted string. Doing so interprets the contents of the string as units and applies them to whatever is in the x register. For example:
0: 100MHz 2pi* "rads/s"
628.32 Mrads/s: 2pi / "Hz"
100 MHz:
0: 9.8066 "m/s^2"
9.8066 m/s^2:
Normally units are given after the number, however a dollar sign would be given immediately before:
0: $100M
$100M:
You can enter hexadecimal, octal, or binary numbers, in either traditional programmers notation or in Verilog notation. For example:
0: 0xFF
255: 0o77
63: 0b1111
15: 'hFF
255: 'o77
63: 'b1111
15:
You can also display numbers in hexadecimal, octal, or binary in both
traditional or Verilog notation. To do so, use hex, oct, bin,
vhex, voct, or vbin:
0: 255
255: hex4
0x00ff: vbin
'b11111111:
You can convert voltages into dBm using:
0: 10 vdbm
30:
You can convert dBm into voltage using:
0: -10 dbmv
100 mV:
Both of these assume a load resistance that is contained in memory Rref, which by default is 50 Ohms.
At start up EC reads and executes commands from files. It first tries ‘~/.ecrc’ and runs any commands it contains if it exists. It then tries ‘./.ecrc’ if it exists. Finally it runs any files given on the command line. It is common to put your generic preferences in ‘~/.exrc’. For example, if your are a physicist with a desire for high precision results, you might use:
eng6
h 2pi / "J-s" =hbar
This tells EC to use 6 digits of resolution and predefines hbar as a constant. The local start up file (‘./.ecrc’) or the file given as a command line argument is generally used to give more project specific initializations. For example, in a directory where you are working on a PLL design you might have an ‘./.ecrc’ file with the following contents:
88.3uSiemens =kdet
9.1G "Hz/V" =kvco
2 =m
8 =n
1.4pF =cs
59.7pF =cp
2.2kOhms =rz
EC also takes commands from the command line. For example:
$ ec "125mV 67uV / db"
65.417
EC prints back-quoted strings while interpolating the values of registers and variables when requested. For example:
$ ec 'degs 500 1000 rtop "V/V" `Gain = $0 @ $1.` quit'
Gain = 1.118 KV/V @ 26.565 degs.
Normally ec prints the value of the x register and exits when it runs out of things to do. The quit at the end tells ec to exit immediately. In this way the value of the x register is not printed. Without it you would see the magnitude printed twice.
You can define functions with the following syntax: ( … )name, where ‘(’ starts the function definition, ‘)name’ terminates it, and … is simply a collection of calculator actions. For example:
0: (2pi * "rads/s")to_omega
0: (2pi / "Hz")to_freq
0: 1.4GHz
1.4 GHz: to_omega
8.7965 Grads/s: to_freq
1.4 GHz:
You can get a list of the actions available with:
0: ?
You can get help on a specific topic, such as //, with:
0: ?//
You can get a list of the help topics available with:
0: help
There is much more available that what is described here. If you have installed the man-page, you can get more information by running:
$ man ec
Alternately, you can view the online documentation.
You can quit the program using:
0: quit
(or :q or ^D).
More detailed information can be found here.