Python : %run to import another folder's ipynb files

This post shows how to use %run command for importing Jupyter Notebook (ipynb) files from another folders. It is essentially to run the imported ipynb files before running a main file.

%run to Import another folder's ipynb files

• Python : Importing an ipynb file (Jupyter Notebook) from another ipynb file

This post is similar to the previous post but easily extended to ipynb files from other folders. So I introduce it to you. It is so easy.

To import ipynb files in another folders, we can use %run command.

As an example, I split one file into three files: a0_load_lib_func.ipynb, a1_read_data.ipynb, and a2_run_rnn.ipynb. The first two files are saved at D:/blog_temp/ for an illustration.

1) Common code block : a1_load_lib_func.ipynb

a1_load_lib_func.ipynb contains package libraries, some user-defined functions.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

# Import Library # In[1]: _________________________________________________________________ import pandas as pd import numpy as np import matplotlib.pyplot as plt from keras.models import Sequential from keras.layers import Dense, SimpleRNN get_ipython().run_line_magic('matplotlib', 'inline') from IPython.core.interactiveshell import InteractiveShell InteractiveShell.ast_node_interactivity = "all" _________________________________________________________________   # Functions # In[2]: _________________________________________________________________ # convert into dataset matrix def convertToMatrix(data, step):     X, Y =[], []     for i in range(len(data)-step):         d=i+step; X.append(data[i:d,]); Y.append(data[d,])     return np.array(X), np.array(Y)   def draw_plot1(df,predicted):     index = df.index.values     plt.figure(figsize=(5, 2.5))     plt.plot(index,df); plt.plot(index,predicted)     plt.show()     return plt _________________________________________________________________   Colored by Color Scripter

cs

2) Reading data : a2_read_data.ipynb

a2_read_data.ipynb is used to read data.

1 2 3 4 5 6 7 8 9 10 11 12 13

# Read Dataset # In[3]: _________________________________________________________________ step = 4; N = 1000; Tp = 800     t=np.arange(0,N) x=np.sin(0.02*t)+2*np.random.rand(N) df = pd.DataFrame(x) # df.head(); plt.plot(df); plt.show() train=df.values train = np.append(train,np.repeat(train[-1,],step)) trainX,trainY = convertToMatrix(train,step) trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1])) _________________________________________________________________

cs

3) Main file : a3_run_rnn.ipynb

The following main file (a3_run_rnn.ipynb) imports two above files and run a SimpleRNN tensorflow function.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

# Import other folder's ipynb files # In[1]: _________________________________________________________________ # the same folder #%run "a1_load_lib_func.ipynb" #%run "a2_read_data.ipynb"   # different folder %run "D:/blog_temp/a1_load_lib_func.ipynb" %run "D:/blog_temp/a2_read_data.ipynb" _________________________________________________________________   # ## Building Model # In[4]: _________________________________________________________________ model = Sequential() model.add(SimpleRNN(units=32, input_shape=(1,step), activation="relu")) model.add(Dense(8, activation="relu"))  model.add(Dense(1)) model.compile(loss='mean_squared_error', optimizer='rmsprop') model.summary() _________________________________________________________________ Model: "sequential" -----------------------------------------------------------------  Layer (type)                Output Shape              Param #    =================================================================  simple_rnn (SimpleRNN)      (None, 32)                1184                                                                          dense (Dense)               (None, 8)                 264                                                                           dense_1 (Dense)             (None, 1)                 9                                                                            ================================================================= Total params: 1,457 Trainable params: 1,457 Non-trainable params: 0 ----------------------------------------------------------------- # ## Training Model # In[5]: _________________________________________________________________ model.fit(trainX, trainY, epochs=100, batch_size=16, verbose=0) _________________________________________________________________ <keras.callbacks.History at 0x22e182f8c70> # In[6]: _________________________________________________________________ trainPredict = model.predict(trainX) trainScore = model.evaluate(trainX, trainY, verbose=0) print(trainScore) draw_plot1(df,trainPredict) _________________________________________________________________ 0.3513454794883728     Colored by Color Scripter

cs