#----------------------------------------------------------- # Lecture 8 -- File IO and Plotting #----------------------------------------------------------- # additional references: # * Numpy savetxt: https://docs.scipy.org/doc/numpy/reference/generated/numpy.savetxt.html # * Numpy loadtxt: https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.loadtxt.html # * Matplotlib documentation -- https://matplotlib.org/contents.html # * Introductory matplotlib tutorial -- https://matplotlib.org/tutorials/introductory/pyplot.html#sphx-glr-tutorials-introductory-pyplot-py # * 10 rules for better figures -- http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1003833 # * matplotlib.pyplot.plot: https://matplotlib.org/api/_as_gen/matplotlib.pyplot.plot.html # * Many examples! -- https://matplotlib.org/gallery/index.html # * More about matplotlib formatting options -- https://matplotlib.org/api/_as_gen/matplotlib.pyplot.plot.html # * More about global figure options (rcParams) -- https://matplotlib.org/devdocs/api/_as_gen/matplotlib.pyplot.rc.html import numpy as np #%% #----------------------------------------------------------- # A. Reading text files using Numpy #----------------------------------------------------------- # # You can read in a data from a text file using the function: # np.loadtxt( , ) # # Comments: # * is a string of a file in your local directory # * You can navigate to a directory in Spyder using the Folder icon in the # top right corner. # * is optional. The default value is any whitespace. # * There are even more options. See the online documentation mydata = np.loadtxt("pipe_data.csv", delimiter=',') print(mydata) print(mydata.shape) # ---------- # Practice # ---------- # Load the file "rate_data.csv" that you saved using Excel # ** Hint use the argument: skiprows=1 to skip the text header #%% #----------------------------------------------------------- # B. Writing text files using Numpy #----------------------------------------------------------- # # Text files can be written using the function: # np.savetxt( , , , ) # # Comments: # * must a string. It will save the file in the local directory. # * must be a variable or numpy array (i.e. vector or matrix) # * is optional. "%.18e" is the default. See online ref for more about this. # * is optional. The default value is a space:' '. # * There are even more options. See the online documentation. t = np.linspace(0, 5, 201) # t (min) np.savetxt("t.dat", t) #%% # Two or more arrays can be combined using np.vstack (an array function # in numpy) and written to the same file. # # * Use the np.vstack function to make a 2-D array out of this data. # * This makes a matrix with a column of t and a column of c. # * Write this data to a text file using np.savetxt t = np.linspace(0,1,10) c = 4*np.exp(-t/1.2)*( 1 + 0.35*np.sin(np.pi*t) ) # c (mM) data = np.vstack([t, c]).T # .T is a transpose print("The rows and cols of data are", np.shape(data)) np.savetxt("rate_data.dat", data) # ---------- # Practice # ---------- # Save x and y to a file comma-delimited called "Gaussian.csv" x = np.linspace(-3, 3, 201) y = 1./np.sqrt(2*np.pi) * np.exp(-x**2/2) #%% #----------------------------------------------------------- # C. Plotting using the matplotlib module #----------------------------------------------------------- # # Matplotlib is a module for making easy, professional quality plots and # figures. It is very powerful and has *many* options. One can make much nicer # figures using matplotlib than using Excel. It is also much easier to save # figures to different formats using matplotlib. import matplotlib.pyplot as plt # ** Menu option to get interactive plots ** # Tools -> Preferences -> IPython console -> Graphics (tab) -> # Graphics Backend, Backend: (drop down menu) -> Automatic # * Then restart Spyder # * Default option is "Inline" -> appears in ipython console # OR can use ipython's "magic" command # from IPython import get_ipython # and either one: # get_ipython().magic('matplotlib inline') # == %matplotlib inline # get_ipython().magic('matplotlib auto') # == %matplotlib auto # A basic plot can be produced by calling the plot( , ) function x = np.linspace(0,6,51) y = np.exp(-x)*np.sin(4*x) plt.plot(x,y) #%% # Some basic formatting # * linestyle: '-', '--', ':', '-.' # * symbols: 'o', 'x', 's', etc. # * colors: 'r', 'g', 'b', 'm', etc. # * labels, title and legend plt.plot(x,y, 'ko-') # linestyle, symbols, colors plt.xlabel('The x-axis label') plt.ylabel('The y-axis label') plt.title('Some title') plt.legend(['my curve']) # ---------- # Practice # ---------- # Load rate_data.csv and plot t vs c. # Label the x- and y-axes as 't (min)' and 'c (mM)' respectively. #%% #----------------------------------------------------------- # D. Examples using more formatting Options #----------------------------------------------------------- # # * Multiple curves on the same plot # * xlim, ylim -- limits of plot # * xscale, yscale -- 'log' or 'linear' x = np.logspace(-2,4,51) y1 = x**(-3) y2 = x**(-0.2) plt.figure() # creates a new figure if one is already open plt.plot(x, y1, 'r--') plt.plot(x, y2, 'bx') plt.xscale('log') plt.yscale('log') plt.xlim([1e-2,1e4]) plt.ylim([1e-2,1e4]) plt.xlabel('x') plt.ylabel('y') plt.legend(['curve 1', 'curve2']); plt.title('log-log plot') plt.show() # forces python to show plot here, instead of waiting until end #%% # * Multiple curves on the same plot # * plot options: color, linewidth, markersize # * global font size: rc() (also has other global parameters) # * legend options: frameon, location, fontsize # * save to file x = np.linspace(0,10,51) y1 = 2*np.exp(-x/2) y2 = 2*np.exp(-x/2)*np.cos(2*x) plt.figure() plt.rc("font", size=10) # Set all font sizes plt.rc('figure', figsize=(8, 6)) # size in inches plt.plot(x, y1, linestyle='--', color='green', linewidth=3) plt.plot(x, y2, linestyle='-', marker='s' , color='blue', markersize=2) plt.xlim([0,10]) plt.ylim([-2,2]) plt.xlabel(r'With Symbols: $\alpha_{s}\beta^{s}$', fontsize=20) # r'' is a string literal plt.ylabel('some label', fontsize=20) plt.legend(['curve 1', 'curve2'], fontsize=16, frameon=False, loc='upper right') plt.tight_layout() # fixes spacing issues around edge of the box # be careful of working directory (make explicit or # use toolbar to navigate to local directory) plt.savefig('saved_figure.png')