new features added

.gitignore

@@ -7,4 +7,4 @@
 *.tif
 build/
 dist/
-*.egg_info/
+*.egg-info/

README.md

@@ -1,34 +1,109 @@
-Bsisb (Xi-cam PluginMaker)
-===============================
+BSISB (A Xi-cam plugin for FTIR data analysis)
+==============================================
 
-version number: 0.1
-author: Liang Chen
+Version number: 0.1
 
-Overview
---------
+Author: Liang Chen
 
+## Installation
+---------------------
+### Install C++ build tools (Windows OS only):
 
+Download and install "Build Tools for Visual Studio" from [here](https://visualstudio.microsoft.com/downloads/).
+See the screenshot below for reference:
 
-Installation / Usage
---------------------
-Install Xi-CAM:
+![build](images/buildtools.png)
 
-Follow the guide [here](https://xi-cam2.readthedocs.io/en/latest/install.html)
+### Install git and python3.7.7:
 
+You will need to ensure that you have both git and python3 installed on your system for Xi-cam installation.
 
-Clone the repo:
+**MacOS**
+
+Open the Terminal application (in Applications/Utilities). In the terminal, check to see if git is installed by typing `git --version`. Either a version number will be printed, indicating git is already installed, or a dialog will open asking The “git” command requires the command line developer tools. Would you like to install the tools now? Click the **Install** button to install the developer tools, which will install git for you.
+
+You will also need to install python3.7.7. You can download python3.7.7 at [python.org](https://www.python.org/downloads/mac-osx/). You will want to get the macOS 64-bit installer if you are running any macOS version since Mavericks.
+
+**Windows OS**
+
+Download git [here](https://git-scm.com/download/win) and follow the installer’s instructions. This will install **Git for Windows**, which provides a **Git Bash** command line interface as well as a **Git GUI** graphical interface for git.
+
+You will want to go to the python3.7.7 download page at [python.org](https://www.python.org/downloads/release/python-377/). For modern systems, install the **Windows x86-64 executable installer** at the bottom.
+
+When you run the installer, make sure sure to check the box that says **Add Python 3.x to PATH** to ensure that the interpreter will be placed in your execution path.
+
+### Create and Activate a Virtual Environment:
+
+The latest python3 version comes with the venv module, which can be used to create a virtual environment. A virtual environment is a sequestered space where you can install and uninstall packages without modifying your system’s installed packages.
+
+Create a virtual environment for installing the Xi-cam components and dependencies. You will then want to activate the virtual environment you created so that any packages you install with python’s package manager, **pip**, will be installed into that active virtual environment. In the commands below, create a virtual environment called **venv_xicam** and activate it:
+
+**Linux/macOS**
+
+    $ python3 -m venv venv_xicam
+    $ source venv_xicam/bin/actviate
+
+**Windows OS**
+
+    $ python -m venv venv_xicam
+    $ venv_xicam\Scripts\activate 
+
+### Install Xi-CAM:
+Before installing Xi-cam, run the following commands:
+
+    $ python -m pip install --upgrade pip
+    $ pip install wheel
+    $ pip install --upgrade setuptools
+    $ pip install numcodecs
+
+Then run the following commands to install Xi-cam:
+
+    $ pip install xicam
+
+### Install lbl_ir and Xi-cam.BSISB plugin:
+Run the following commands to install lbl_ir package and Xi-cam.BSISB plugin:
 
     $ git clone https://github.com/lchen23/Xi-cam.BSISB.git
-    $ cd Xi-CAM.BSISB
+    $ cd Xi-cam.BSISB/lbl_ir
     $ pip install -e .
-    $ cd .. 
+    $ cd ..
+    $ pip install -e .
+
+### Run the Xi-cam.BSISB program:
+
+Run the following command:
+
     $ xicam
+
+### Run the Xi-cam.BSISB GUI from a new terminal:
+First, activate the virtual environment, then run xicam:
+
+**Linux/macOS**
+
+    $ source venv_xicam/bin/actviate
+    $ xicam
+
+**Windows OS**
+
+In Git bash:
+
+    $ source venv_xicam/Scripts/activate
+    $ xicam 
+
+In Windows CMD:
+
+    $ venv_xicam\Scripts\activate
+    $ xicam 
+
+After the Xi-cam GUI program launches, open the
+`Xi-cam.BSISB/tests/PC12-NGF-3h.h5` file by double clicking the h5 file
+in the file browser as shown below:
+
+![example](images/example.png)
+
 Contributing
 ------------
 
 TBD
 
-Example
--------
 
-TBD

lbl_ir/.gitignore

@@ -0,0 +1,2 @@
+__pycache__/
+.ipynb_checkpoints/

lbl_ir/README.md

@@ -0,0 +1,30 @@
+# README #
+
+
+This README would normally document whatever steps are necessary to get your application up and running.
+
+### What is this repository for? ###
+
+* Quick summary
+* Version
+* [Learn Markdown](https://bitbucket.org/tutorials/markdowndemo)
+
+### How do I get set up? ###
+
+* Summary of set up
+* Configuration
+* Dependencies
+* Database configuration
+* How to run tests
+* Deployment instructions
+
+### Contribution guidelines ###
+
+* Writing tests
+* Code review
+* Other guidelines
+
+### Who do I talk to? ###
+
+* Repo owner or admin
+* Other community or team contact

lbl_ir/lbl_ir/GPR/GPR_engine.py

@@ -0,0 +1,143 @@
+import numpy as np
+import matplotlib.pyplot as plt
+import numba
+from scipy.optimize import minimize
+from scipy.optimize import basinhopping
+
+@numba.jit(nopython=True) #,cache=True)
+def dmat(X1,X2=None):
+    if X2 is None:
+        X2 = X1
+    N1,M1 = X1.shape
+    N2,M2 = X2.shape
+
+    dd = np.zeros( (N2,N1) )
+    for ii in range(N2):
+        for jj in range(N1):
+            tmp = 0
+            for kk in range(M1):
+                delta = X2[ii,kk]-X1[jj,kk]
+                tmp += delta*delta
+            tmp = np.sqrt( tmp )
+            dd[ii,jj]=tmp
+    return dd
+
+@numba.jit(nopython=True) #,cache=True)
+def rbf_kern(dmat, variance, length):
+    result = np.exp( -dmat*dmat/(1e-8+length*length*2.0))*(1e-8+variance)
+    return result
+
+@numba.jit(nopython=True) #,cache=True)
+def d_rbf_kern_FD(X2, X, variance, length, h=1e-4):
+    dmat_plus = dmat(X,X2+h/2)
+    dmat_minus= dmat(X,X2-h/2)
+    Kplus     = np.exp( -dmat_plus*dmat_plus/(length*length*2.0) )*variance
+    Kminus    = np.exp( -dmat_minus*dmat_minus/(length*length*2.0) )*variance
+    result    = (Kplus - Kminus)/h
+    return result
+
+def d_rbf_kern(X2, X, dxx2, Kxx2, variance, length):
+    # first derivative
+    result1 = []
+    for xx in X2.flatten():
+        result1.append( X.flatten()-xx )
+    result1 = np.vstack(result1)
+    result1 = result1*Kxx2
+
+    # second derivative
+    result2 = []
+    N2,N1 = result1.shape
+    for ii in range(N2):
+        xx = X2[ii,:]
+        dK = result1[ii,:]
+        K  = Kxx2[ii,:]
+        tmp = -K/(length*length)
+        tmp2 = (X-xx).flatten()*dK.flatten()/(length*length)
+        result2.append( tmp.flatten()+tmp2.flatten() )
+    result2 = np.vstack(result2)
+    return result1, result2
+
+
+
+class GPR_exp_fitter(object):
+    def __init__(self,sigma,length, mu=0,niter=3):
+        """
+        :param sigma: The sigma
+        :param length: The length scale
+        :param mu: The mean value
+        :param niter: the number of iterations in basin hopping to fit the GPR model
+        """
+
+        self.sigma = sigma
+        self.length = length
+        self.mu = mu
+
+        # to compute later
+        self.dX1X1 = None
+
+
+    def marginal_likelihood(self,hparams):
+        this_mu    = hparams[0]
+        this_sigma = hparams[1]
+        this_length= hparams[2]
+        KX1X1      = rbf_kern( self.dX1X1, this_sigma*this_sigma, this_length ) 
+
+        II         = np.diag(self.sY*self.sY)
+        KX1X1_inv  = np.linalg.pinv(KX1X1+II)
+        tY         = self.Y - this_mu
+        tY         = tY.reshape(-1,1)
+        KiY        = KX1X1_inv.dot(tY)
+        term1      = 0.5*tY.transpose().dot(KiY)    
+        term2      = 0.5*np.linalg.slogdet(KX1X1+II)[1]
+
+        result     = term1+term2
+        result     = result.flatten()[0]
+        return result
+
+
+    def fit(self,X,Y,sY):
+        # first we need to build a distance matrix
+        self.sY = sY
+        if type(self.sY) is float:
+            self.sY = self.Y*0+sY
+        self.X = X
+        self.Y = Y
+        self.dX1X1 = dmat(X.reshape(-1,1))
+        fitter = basinhopping( func   = self.marginal_likelihood,
+                               x0     = np.array([self.mu, self.sigma, self.length]),
+                               niter  = 13)
+        self.mu      = fitter.x[0]
+        self.sigma   = abs(fitter.x[1])
+        self.length  = abs(fitter.x[2])
+
+
+        self.tY    = (Y - self.mu).reshape(-1,1)
+        self.KX1X1 = rbf_kern( self.dX1X1, self.sigma**2.0, self.length ) + np.eye( self.dX1X1.shape[0])*self.sY
+        self.KX1X1_inv = np.linalg.pinv(self.KX1X1)
+        self.KiY = self.KX1X1_inv.dot(self.tY)
+
+
+
+    def predict(self,X):
+        dX1X2 = dmat(self.X.reshape(-1,1), X.reshape(-1,1))
+        kX1X2 = rbf_kern( dX1X2, self.sigma**2, self.length )
+        tmp   = kX1X2.dot( self.KiY )
+        return tmp+self.mu  
+
+
+
+
+def tst():
+    x   = np.linspace(-10,10,20, True)
+    xx  = np.linspace(-10,10,200,True)
+    y   = np.sin(x) #10 + x*x + x
+
+    plt.plot(x,y,'.');plt.show()
+    obj = GPR_exp_fitter(1,1,1) 
+    obj.fit(x,y,0.01)
+    yy = obj.predict( xx ).flatten()
+    plt.plot(x,y,'.'); plt.plot(xx,yy);plt.show()
+
+
+if __name__ == "__main__":
+    tst()