{"cells":[{"cell_type":"markdown","id":"e8a1cfdb","metadata":{"id":"e8a1cfdb"},"source":["[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](http://colab.research.google.com/github/ai2es/WAF_ML_Tutorial_Part1/blob/main/colab_notebooks/Notebook07_ComplexMLRegression.ipynb)\n","\n","# Notebook 07: Complex ML Regression [Colab Version]\n","\n","### Goal: Training a ML using all features/predictors/inputs \n","\n","#### Reminder of Problem Statement\n","\n","\n","Reminder of the ML task we want to accomplish in the paper. \n","\n","1. Does this image contain a thunderstorm? <-- Classification\n","2. How many lightning flashes are in this image? <-- Regression\n","\n","#### Background\n","\n","Please make sure you already did Notebook 5, because this notebook extends what we did in Notebook 5 to now include many more input predictors. This is the same thing as Notebook 6, but for the regression task.\n","\n","#### Step 0: Get the github repo (we need some of the functions there)\n","\n","The first step with all of these Google Colab notebooks will be to grab the github repo and cd into the notebooks directory. \n","\n","To run things from the command line, put a ```!``` before your code"]},{"cell_type":"code","source":["#get the github repo \n","!git clone https://github.com/ai2es/WAF_ML_Tutorial_Part1.git \n","\n","#cd into the repo so the paths work \n","import os \n","os.chdir('/content/WAF_ML_Tutorial_Part1/jupyter_notebooks/')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"qzwtyeVAa5RQ","executionInfo":{"status":"ok","timestamp":1648653782183,"user_tz":300,"elapsed":29110,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"ccb9b2ba-aa7a-4b5b-f5db-ac1215457cd1"},"id":"qzwtyeVAa5RQ","execution_count":1,"outputs":[{"output_type":"stream","name":"stdout","text":["Cloning into 'WAF_ML_Tutorial_Part1'...\n","remote: Enumerating objects: 301, done.\u001b[K\n","remote: Counting objects: 100% (301/301), done.\u001b[K\n","remote: Compressing objects: 100% (197/197), done.\u001b[K\n","remote: Total 301 (delta 139), reused 236 (delta 96), pack-reused 0\u001b[K\n","Receiving objects: 100% (301/301), 195.77 MiB | 13.14 MiB/s, done.\n","Resolving deltas: 100% (139/139), done.\n","Checking out files: 100% (100/100), done.\n"]}]},{"cell_type":"markdown","source":["#### Step 1 & 2: Import packages and load data for Classification \n","In Notebook 5 we only wanted 1 feature, no we want all available inputs (36 total). So all we need to change is the ```features_to_keep``` keyword to include all indices. Remember in notebook 5 we wanted to drop the zeros to help. lets do that right away"],"metadata":{"id":"TzWno3Jga3Iq"},"id":"TzWno3Jga3Iq"},{"cell_type":"code","execution_count":2,"id":"eed55e53","metadata":{"id":"eed55e53","executionInfo":{"status":"ok","timestamp":1648653807513,"user_tz":300,"elapsed":8642,"user":{"displayName":"Randy C","userId":"12301342565919601334"}}},"outputs":[],"source":["#needed packages \n","import xarray as xr\n","import matplotlib.pyplot as plt \n","import numpy as np\n","import pandas as pd\n","\n","#plot parameters that I personally like, feel free to make these your own.\n","import matplotlib\n","matplotlib.rcParams['axes.facecolor'] = [0.9,0.9,0.9] #makes a grey background to the axis face\n","matplotlib.rcParams['axes.labelsize'] = 14 #fontsize in pts\n","matplotlib.rcParams['axes.titlesize'] = 14 \n","matplotlib.rcParams['xtick.labelsize'] = 12 \n","matplotlib.rcParams['ytick.labelsize'] = 12 \n","matplotlib.rcParams['legend.fontsize'] = 12 \n","matplotlib.rcParams['legend.facecolor'] = 'w' \n","matplotlib.rcParams['savefig.transparent'] = False\n","\n","#make default resolution of figures much higher (i.e., High definition)\n","%config InlineBackend.figure_format = 'retina'\n","\n","#import some helper functions for our other directory.\n","import sys\n","sys.path.insert(1, '../scripts/')\n","from aux_functions import load_n_combine_df\n","(X_train,y_train),(X_validate,y_validate),(X_test,y_test) = load_n_combine_df(path_to_data='../datasets/sevir/',features_to_keep=np.arange(0,36,1),class_labels=False,dropzeros=True)"]},{"cell_type":"markdown","id":"4f8089e4","metadata":{"id":"4f8089e4"},"source":["I have a habit of always checking shapes, so let's do that "]},{"cell_type":"code","execution_count":4,"id":"6abf2270","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"6abf2270","executionInfo":{"status":"ok","timestamp":1648653819556,"user_tz":300,"elapsed":373,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"6ab75757-6d44-486b-c3d1-42ef7c1eb586"},"outputs":[{"output_type":"stream","name":"stdout","text":["X_train, y_train shapes: (222051, 36),(222051,)\n","X_val, y_val shapes: (47319, 36),(47319,)\n","X_test, y_test shapes: (47376, 36),(47376,)\n"]}],"source":["print('X_train, y_train shapes: {},{}'.format(X_train.shape,y_train.shape))\n","print('X_val, y_val shapes: {},{}'.format(X_validate.shape,y_validate.shape))\n","print('X_test, y_test shapes: {},{}'.format(X_test.shape,y_test.shape))"]},{"cell_type":"markdown","id":"d48b8b07","metadata":{"id":"d48b8b07"},"source":["Good, these shapes are indeed less than the classification task (Notebook 6) because we are dropping the zeros.\n","\n","#### Change from Notebook 5\n","\n","Since we are using more than 1 input predictor it is important to normalize our predictors. Why is this important? because in reality each one of our inputs have a range of valid values associated with them, and that range could be large (e.g., -100 to 100) or it could be small (e.g., 0,1). The machine learning will weight these inputs quantitatively, so if we use the default scalings, it might be biased to use the larger magnitude predictors more than the small magnitude predictors. To prevent this we will scale the data to have mean 0, and variance 1. You are likely more familiar with the term *standard anomaly* which is the same thing: \n","\n","$$ z = \\frac{x - \\mu}{\\sigma} $$\n","\n","where $\\mu$ is the mean of that specific feature and $\\sigma$ is the standard deviation, both calculated from the training dataset. We could implement this ourselves, but ```sklearn``` has this built for us: ```sklearn.preprocessing.StandardScaler``` and it works alot like how we fit the machine learning model before."]},{"cell_type":"code","execution_count":5,"id":"90afa794","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"90afa794","executionInfo":{"status":"ok","timestamp":1648653827130,"user_tz":300,"elapsed":781,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"98e992cf-21e8-4763-d6b9-3e7737d9cf86"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["(array([ 7.74633690e-16, -1.56667488e-16, -2.87735713e-16, -2.57528583e-16,\n"," 2.30393364e-17, 2.35513217e-17, -6.94380001e-17, -7.21899209e-17,\n"," 1.77530887e-16, 2.40223481e-15, -6.88108182e-16, -4.63858640e-16,\n"," -6.53293185e-16, 3.88852801e-16, -4.31987558e-16, -2.67256303e-16,\n"," 1.10550416e-15, -2.05306087e-16, 1.10076830e-17, 2.81591890e-18,\n"," -3.83988941e-17, -1.47195761e-16, -1.43355871e-17, -4.91505844e-17,\n"," -3.99348498e-17, 9.62532278e-17, 1.97370316e-16, -2.45752922e-17,\n"," -2.96951448e-17, 1.04956977e-17, -3.96788572e-18, 4.09588204e-18,\n"," -3.59029660e-17, -3.83988941e-19, 3.21270747e-17, 3.47637988e-16]),\n"," array([1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n"," 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n"," 1., 1.]))"]},"metadata":{},"execution_count":5}],"source":["from sklearn.preprocessing import StandardScaler\n","\n","#create scaling object \n","scaler = StandardScaler()\n","#fit scaler to training data\n","scaler.fit(X_train)\n","\n","#transform feature data into scaled space \n","X_train = scaler.transform(X_train)\n","X_validate = scaler.transform(X_validate)\n","X_test = scaler.transform(X_test)\n","\n","#double check that mean is 0 and std is 1. \n","np.mean(X_train,axis=0),np.std(X_train,axis=0)"]},{"cell_type":"markdown","id":"807c6b96","metadata":{"id":"807c6b96"},"source":["Note that e-16 means $\\times 10^{-16}$ which is effectively 0. So it has successfully scaled our data to now have mean 0 and std 1. We are now ready to train our machine learning model again."]},{"cell_type":"markdown","id":"9b49b904","metadata":{"id":"9b49b904"},"source":["#### Step 3: Initialize model\n","\n","To keep consistency with Notebook 5, we will continue using linear regression"]},{"cell_type":"code","execution_count":6,"id":"f200b4ba","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"f200b4ba","executionInfo":{"status":"ok","timestamp":1648653830769,"user_tz":300,"elapsed":425,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"06498ae7-4ea8-4758-8402-6e678a4af3d3"},"outputs":[{"output_type":"stream","name":"stdout","text":["LinearRegression()\n"]}],"source":["#load model from sklearn\n","from sklearn.linear_model import LinearRegression\n","\n","#initialize\n","model = LinearRegression()\n","\n","print(model)\n"]},{"cell_type":"markdown","id":"bd0f6ae0","metadata":{"id":"bd0f6ae0"},"source":["#### Step 4: Train your ML model! "]},{"cell_type":"code","execution_count":7,"id":"1aa1af41","metadata":{"id":"1aa1af41","executionInfo":{"status":"ok","timestamp":1648653833158,"user_tz":300,"elapsed":786,"user":{"displayName":"Randy C","userId":"12301342565919601334"}}},"outputs":[],"source":["model = model.fit(X_train,y_train)"]},{"cell_type":"markdown","id":"7292658d","metadata":{"id":"7292658d"},"source":["#### Step 5: Evaluate your ML model\n","\n","As a sanity check, we will first look at the *one-to-one* plot where the x-axis is the predicted number of flashes, and the y-axis is the true number of flashes. A perfect prediction will be directly along the diagonal. "]},{"cell_type":"code","execution_count":8,"id":"8d391c52","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":359},"id":"8d391c52","executionInfo":{"status":"ok","timestamp":1648653835838,"user_tz":300,"elapsed":953,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"42ae3dc4-4711-48dc-8173-b0d19180cfec"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["Text(0.5, 0, 'GLM measurement, [$number of flashes$]')"]},"metadata":{},"execution_count":8},{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"image/png":{"width":331,"height":324}}}],"source":["#get predictions \n","yhat = model.predict(X_validate)\n","\n","#color I like. The order of ratios is [Red,Green,Blue]\n","r = [255/255,127/255,127/255]\n","\n","#make figure \n","fig = plt.figure(figsize=(5,5))\n","#set background color to white so we can copy paste out of the notebook if we want \n","fig.set_facecolor('w')\n","\n","#get axis for drawing\n","ax = plt.gca()\n","\n","#plot data \n","ax.scatter(yhat,y_validate,color=r,s=1,marker='+')\n","ax.plot([0,3500],[0,3500],'-k')\n","ax.set_xlabel('ML Prediction, [$number of flashes$]')\n","ax.set_xlabel('GLM measurement, [$number of flashes$]')"]},{"cell_type":"markdown","id":"44c142b7","metadata":{"id":"44c142b7"},"source":["To me this looks better than the 1 input feature model we showed in Notebook 5. One troubling thing with these types of scatter plots though is that there are so many points, it is hard to see where the 'mode' of the distribution is. To look a bit deeper into this plot, and to match the plots we used in the paper (Figures 14 and 16). I have a function that will go ahead and bin up the data on this plot and count how many points are in each bin. This function is called ```boxbin``` and it is located in the ```aux_functions.py``` script"]},{"cell_type":"code","execution_count":9,"id":"f48575cc","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":382},"id":"f48575cc","executionInfo":{"status":"ok","timestamp":1648653842966,"user_tz":300,"elapsed":1866,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"11de3d4d-7156-4e6b-e1ea-43e504a62a4d"},"outputs":[{"output_type":"stream","name":"stdout","text":["n_samples= 27668.0\n"]},{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"image/png":{"width":351,"height":347}}}],"source":["from aux_functions import boxbin\n","\n","#make figure \n","fig = plt.figure(figsize=(5,5))\n","#set background color to white so we can copy paste out of the notebook if we want \n","fig.set_facecolor('w')\n","\n","#get axis for drawing\n","ax = plt.gca()\n","\n","#build the bins we want to use for our data\n","n = 33\n","xbins = np.logspace(0,3.5,n) #this logspace logarithmically spaces bins from 10^0 to 10^3.5. The lightning data is better shown on a log-log scale\n","ybins = np.logspace(0,3.5,n) #we choose to keep the bins square \n","\n","#Color I like\n","r = [255/255,127/255,127/255]\n","\n","#make axes log-log \n","ax.semilogy()\n","ax.semilogx()\n","#same scatter as before \n","ax.scatter(yhat,y_validate,color=r,s=1,marker='+')\n","\n","#use the boxbin function to bin data!\n","ax,cbar,C = boxbin(yhat,y_validate,xbins,ybins,ax=ax,mincnt=100,normed=True,cmap='Reds_r',vmin=0,vmax=2)\n","\n","cbar.set_label('$\\%$ of all points')\n","ax.set_xlim([1,4000])\n","ax.set_xticks([1,10,100,1000])\n","ax.set_yticks([1,10,100,1000])\n","ax.set_ylim([1,4000])\n","ax.plot([1,4000],[1,4000],'--k',alpha=0.5)\n"," \n","ax.set_ylabel('$y$, [# of flashes]')\n","ax.set_xlabel(r'$\\hat{y}$, [# of flashes]')\n"," \n"," \n","plt.tight_layout()"]},{"cell_type":"markdown","id":"6c36afe7","metadata":{"id":"6c36afe7"},"source":["Ah that’s better, we can now see where the highest density of points are. Note that only bins with more than 100 points in them will be colored in the boxes. \n","\n","Let's check on the quantitative metrics to see if it is indeed working better with more features. We will calculate these: \n","\n","$$ \\mathrm{Bias} = \\frac{1}{N} \\sum_{j=1}^{N} (y_j - \\hat{y}_j) $$\n","\n","$$ \\mathrm{MAE} = \\frac{1}{N} \\sum_{j=1}^{N} |y_j - \\hat{y}_j| $$\n","\n","$$ \\mathrm{RMSE} = \\sqrt{\\frac{1}{N} \\sum_{j=1}^{N} (y_j - \\hat{y}_j)^{2}} $$ \n","\n","$$ \\mathrm{R^{2}} = 1 - \\frac{\\sum_{j=1}^{N} (y_j - \\hat{y}_j)^{2}}{\\sum_{j=1}^{N} (y_j - \\bar{y})^{2}} $$\n","\n","All of these metrics again in the ```gewitter_functions.py``` script."]},{"cell_type":"code","execution_count":10,"id":"d4a97374","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"d4a97374","executionInfo":{"status":"ok","timestamp":1648653848701,"user_tz":300,"elapsed":362,"user":{"displayName":"Randy C","userId":"12301342565919601334"}},"outputId":"221f1a88-126f-47ee-d2c8-98fd5d1ef9ad"},"outputs":[{"output_type":"stream","name":"stdout","text":["MAE:97.6 flashes, RMSE:174.14 flashes, Bias:31.88 flashes, Rsquared:0.56\n"]}],"source":["from gewitter_functions import get_mae,get_rmse,get_bias,get_r2\n","\n","yhat = model.predict(X_validate)\n","mae = get_mae(y_validate,yhat)\n","rmse = get_rmse(y_validate,yhat)\n","bias = get_bias(y_validate,yhat)\n","r2 = get_r2(y_validate,yhat)\n","\n","#print them out so we can see them \n","print('MAE:{} flashes, RMSE:{} flashes, Bias:{} flashes, Rsquared:{}'.format(np.round(mae,2),np.round(rmse,2),np.round(bias,2),np.round(r2,2)))"]},{"cell_type":"markdown","id":"6e45ffa1","metadata":{"id":"6e45ffa1"},"source":["The numbers from the single feature model are: *MAE:146.68 flashes, RMSE:235.46 flashes, Bias:30.98 flashes, Rsquared:0.2*. So things have indeed improved on the MAE, Bias and Rsquared metrics. But this is a linear model, so trying to use a linear model for a non-linear forecast (like the number of flashes in a satellite image) is not great, but does do ok. "]},{"cell_type":"code","execution_count":null,"metadata":{"id":"YO8nGm4Basq_"},"outputs":[],"source":[""],"id":"YO8nGm4Basq_"}],"metadata":{"kernelspec":{"display_name":"Python 3 (ipykernel)","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.9.7"},"colab":{"name":"Notebook07_ComplexMLRegression.ipynb","provenance":[]}},"nbformat":4,"nbformat_minor":5}