itom cheat sheet Common Syntax Structures Common Functions of Module math (from math import *)

exp [any expression], stmt [(sequence of) command(s)] cos(x), sin(x), tan(x) Cosine, sine, tangent of x radians
Python 3 Note: Indentation is important for control sequences! sqrt(x) Positive square root of x
Assignment a = 1 a=1 degrees(x), radians(x) Convert from rad to deg, deg to rad
Help
a, b = 1, 2 a=1,b=2 exp(x) e ** x
help(m) Display help for module, function…
1 c = [1,2,3]; c[1] = 4 c=[1,4,3] floor(x) Largest whole number <= x
pluginHelp(“name”) Display information for plugin
1 Output print(exp [,expr2…]) print(“test”) pow(x,y) x ** y
filterHelp(“name”) Display information for itom-filter
1 Comment #single line ‘’’multi pi Math constant (15 sig figs)
widgetHelp(“name”) Display information for widget in plugin
line’’’ e Math constant e (15 sig figs)
dir(m) Display names in module m
Selection if(boolean_exp): if(2>1):
stmt print(“2>1”) Common List (L) and Tuple (T) Methods
Common Data Types [elif (boolean_exp): else: LT[idx], LT[idx1:idx2] get items or slice of items from list/tuple
int Integer (32/64bit) 3, -4, 0 stmt] print(“what?”) LT.count(obj) number of occurrences of obj in LT
float Floating point number 3.0, -6.55, float(‘nan’) [else: LT.index(obj) index of first occurrence of obj in LT;
complex Complex number 2+3j, 4j, 5-0j stmt] raises ValueError if does not occur
bool Boolean True, False Repetition while(boolean_exp): repeat while bool_exp L[idx]=obj assigns new value to index (list only)
str String of characters “Python” stmt is True L.append(obj) Appends obj to end of list L
byte Sequence of integers b”Python” L.remove(obj) Removes first occurrence of obj from L
tuple Immutable sequence (2,), (2.3,”a”), (2.3,-1) Traversal for var in obj: Iterate over all
list Mutable sequence [2], [2.3,”a”], [2.3,-1] stmt elements in traversable
Common List (L) or Tuple (T) Methods, (LT both)
dict Mapping, dictionary {“x”:-2, “name”:”a”} obj.
LT[idx], LT[idx1:idx2] get items or slice of items from list/tuple
numpy.array Numpy-Array Loop for i in range(0,5): Use range for creating
LT.count(obj) number of occurrences of obj in LT
dataObject
1
itom data object compatible to np.array print(i) an iterable list
LT.index(obj) index of first occurrence of obj in LT;
[0,1,2,3,4]
raises ValueError if does not occur
Exception try: try:
Module Import L[idx]=obj assigns new value to index (list only)
Handling stmt … 1/0
Example: How to call method plot of module itom L.append(obj) Appends obj to end of list L
except [exc_type] [,var]: except ZeroDivisionError:
import itom itom.plot(args,…) L.remove(obj) Removes first occurrence of obj from L
stmt … print(“uups”)
from itom import plot plot(args,…) Function def fctname(params): def test(i,j=4):
from itom import * plot(args,…) [Import all] Definition ‘’’doc-string’’’ a=i+j Common Dictionary (D) Methods
from itom import plot as fct fct(args,…) [Alias] stmt #j has default 4 D[“key”] returns value corresponding to key
return obj return [a,”done”] D[“key”] = obj replaces/adds obj under given key
Function ret = fctname(args) ret = test(2) “key” in D True if key exists in D, else False
Operators and their Precedence D.clear() clears dictionary
Call #ret is [6,”done”]
func_name(args,kwds) Function call D.keys() Returns list of D’s keys
x[startIdx : endIdx] Slicing (startIdx incl., endIdx excl.) D.values() Returns list of D’s values
x[index] Indexing (index zero-based) Common Built-in Functions
x.attribute Attribute reference abs(x) Absolute value of x Formatting Numbers as Strings
** Exponentation float(x), int(x) Convert x to float / int (if possible) Syntax: “%width.precision type” % expression
*, /, % Multiply, Divide, Mod len(s) Number of items in sequence (list, tuple,...)
width (optional) total width (+/-: right/left aligned)
+, - Add, Subtract str(obj) String representation of obj
precision (optional) specified digits of float precision
&, |, ^, ~ Binary And, Or, Xor, Not range(x,y) A list [x, x+1, x+2, …, y-1] (y excluded)
type (required) d (int), f (float), s (string), e (exp. Notation)
>, <, <=, >=, !=, == Comparison dict() Empty dictionary
Examples: “%6d” % 123 -> …123
in, not in Membership tests [2 in (1,2,3)]->True list() Empty list
“%04d” % 1 -> 0001
not, and, or Boolean operators tuple() Empty tuple
“%8.2f” % 456.789 -> ..456.79
“%8.2e” % 456.789 -> 4.57e+02
1
only available in itom
1 1
Working with dataIO-Devices (Grabber, AD-Converter…) Numpy.array (import numpy as np, np.array), DataObject (import itom, itom.dataObject)
pluginHelp(“name”) Prints information about plugin arr=np.ndarray([2,3],’uint8’) dObj=dataObject([2,3],’uint8’) create a randomly filled 2x3 array with type uint8
dataIO(“name”,params) Creates obj (instance) of device arr=np.array([[1,2,3],[4,5,6]]) dObj =dataObject([2,3],data=(1,2,3,4,5,6)) create the 2x3 array [1,2,3 ; 4,5,6]
obj.getParam(“name”) Returns value of parameter arr=np.array(dObj, copy=False) dObj =dataObject(arr) convert np.array <-> dataObject
obj.setParam(“name”,val) Sets parameter to val arr.ndim dObj.dims Returns number of dimensions (here: 2)
obj.startDevice() Starts device (camera…) arr.shape dObj.shape Returns size tuple (here: [2,3])
obj.stopDevice() Stops device (camera…) arr.shape[0] dObj.size(0) or dObj.shape[0] Returns size of first dimensions (here: y-axis)
st nd
obj.acquire() triggers image acquisition c=arr[0,1]; arr[0,1]=7 dObj [0,1]; b[0,1]=7 Gets or sets the element in the 1 row, 2 col
nd
obj.getVal(dObj) after call, dataObject dObj c=arr[:,1:3] or c=dObj[:,1:3] or Returns shallow copy of array containing the 2 and
rd
references to last acquired image c=arr[0:2,1:3] c= dObj [0:2,1:3] 3 columns
obj.copyVal(dObj) after call, dObj contains deep copy arr[:,:]=7 dObj [:,:]=7 sets all values of array to value 7
of last acquired image arr.transpose() (shallow copy) dObj.trans() (deep copy) transpose of array
obj.setAutoGrabbing(bool) En-/Disables continuous grab for np.dot(arr1,arr2) dObj1 * dObj2 (float only) matrix multiplication
connected live views arr1 * arr2 dObj1.mul(dObj2) element-wise multiply
arr1 / arr2 dObj1.div(dObj2) element-wise divide
1 arr1 +,- arr2 dObj1 +,- dObj2 sum/difference of elements
Working with actuator-Devices (Motors, Stages…)
Position units are in mm arr1 +,- scalar dObj1 +,- scalar adds/subtracts scalar from every element in array
pluginHelp(“name”) Prints information about plugin arr1 &,| arr2 dObj1 &,| dObj2 element-wise, bitwise AND/OR operator
actuator(“name”,params) Creates obj (instance) of device arr2 = arr1 dObj2 = dObj1 referencing (both still point to the same array)
obj.getParam(“name”) Returns value of parameter arr2 = arr1.copy() dObj2 = dObj1.copy() deep copy (entire data is copied)
obj.setParam(“name”,val) Sets parameter to val arr2 = arr1.astype(newtype) dObj2 = dObj1.astype(‘newtypestring’) type conversion
obj.getPos(idx1[,idx2…]) Returns current position for all arr = np.zeros([3,4],’float32’) dObj = dataObject.zeros([3,4], ’float32’) 3x4 array filled with zeros of type float32
given axes indices (0-based) arr = np.ones([3,4],’float32’) dObj = dataObject.ones([3,4], ’float32’) 3x4 array filled with ones of type float32
obj.setPosRel(idx1,pos1,…) Relatively moves axis idx1 by pos1 arr = np.eye(3,dtype=’float32’) dObj = dataObject.eye(3, ’float32’) 3x3 identity matrix (type: float32)
obj.setPosAbs(idx1,pos1,…) Moves axis idx1 to pos1 arr2 = arr1.squeeze() dObj2 = dObj1.squeeze() converts array to an array where dimensions of size
1 are eliminated (deep copy if necessary)
1 np.linspace(1,3,4) - 4 equally spaced samples between 1 and 3, inclusive
Working with itom-Filters
[x,y] = np.meshgrid(0:2,1:5) - two 2D arrays: one of x values, the other of y values
filterHelp(“name”) Lists all algorithms/filters
np.linalg.inv(a) - inverse of square matrix a
containing name or detailed
x=np.linalg.solve(a,b) - solution of ax=b (using pseudo inverse)
information about filter that
[U,S,V] = np.linalg.svd(a) - singular value decomposition of a (V is transposed!)
matches name
np.fft.fft2(a), np.fft.ifft2(a) filter available (Inverse) 2D fourier transform of a
ret=filter(“name”,param1,…) Calls filter name with given
a[a>0]=5 a[a>0]=5 sets all elements > 0 of a to 5
parameters and returns tuple of
a[np.isnan(a)]=5 a[np.isnan(a)]=5 sets all NaN elements of a to 5
output parameters (or None)
arr2 = arr1.reshape([3,2]) arr2 = arr1.reshape([3,2]) reshapes arr1 to new size (equal number of items)
1
Plots
plot(dObj) 1D or 2D plot of dObj Subject Matlab Python/Numpy-Arrays/DataObjects
(depending on its size) Data Copying Matlab always uses deep copying. Python usually creates shallow copies (deep copy only if
liveImage(dataIO-instance) Live view of camera b = a -> b and a contain separated data in memory necessary). Therefore a and b share the same data.
Indexing Matlab uses one-based indexing Python always uses zero-based indexing
1
Common DataObject and Numpy.Array Data Types Ranges 1:4 means the items at one-based indices [1,2,3,4] In Python the same is achieved by 0:4 -> [0,1,2,3]
Both boundaries are included in the range. The second boundary is always excluded!
“uint8”, “int8”, “uint16”, (Un-)Signed integer 8,16,32 bit
“int16”, “uint32”, “int32”
“float32”, “float64” Floating point numbers
“complex64”, “complex128” Complex values (64 = 2x32 bit)
“rgba32” Color value (type itom.rgba)