{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Forecasting, updating datasets, and the \"news\"\n",
    "\n",
    "In this notebook, we describe how to use Statsmodels to compute the impacts of updated or revised datasets on out-of-sample forecasts or in-sample estimates of missing data. We follow the approach of the \"Nowcasting\" literature (see references at the end), by using a state space model to compute the \"news\" and impacts of incoming data.\n",
    "\n",
    "**Note**: this notebook applies to Statsmodels v0.12+. In addition, it only applies to the state space models or related classes, which are: `sm.tsa.statespace.ExponentialSmoothing`, `sm.tsa.arima.ARIMA`, `sm.tsa.SARIMAX`, `sm.tsa.UnobservedComponents`, `sm.tsa.VARMAX`, and `sm.tsa.DynamicFactor`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:06.039932Z",
     "iopub.status.busy": "2022-11-02T17:10:06.039441Z",
     "iopub.status.idle": "2022-11-02T17:10:07.951084Z",
     "shell.execute_reply": "2022-11-02T17:10:07.950275Z"
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import statsmodels.api as sm\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "macrodata = sm.datasets.macrodata.load_pandas().data\n",
    "macrodata.index = pd.period_range('1959Q1', '2009Q3', freq='Q')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Forecasting exercises often start with a fixed set of historical data that is used for model selection and parameter estimation. Then, the fitted selected model (or models) can be used to create out-of-sample forecasts. Most of the time, this is not the end of the story. As new data comes in, you may need to evaluate your forecast errors, possibly update your models, and create updated out-of-sample forecasts. This is sometimes called a \"real-time\" forecasting exercise (by contrast, a pseudo real-time exercise is one in which you simulate this procedure).\n",
    "\n",
    "If all that matters is minimizing some loss function based on forecast errors (like MSE), then when new data comes in you may just want to completely redo model selection, parameter estimation and out-of-sample forecasting, using the updated datapoints. If you do this, your new forecasts will have changed for two reasons:\n",
    "\n",
    "1. You have received new data that gives you new information\n",
    "2. Your forecasting model or the estimated parameters are different\n",
    "\n",
    "In this notebook, we focus on methods for isolating the first effect. The way we do this comes from the so-called \"nowcasting\" literature, and in particular Bańbura, Giannone, and Reichlin (2011), Bańbura and Modugno (2014), and Bańbura et al. (2014). They describe this exercise as computing the \"**news**\", and we follow them in using this language in Statsmodels.\n",
    "\n",
    "These methods are perhaps most useful with multivariate models, since there multiple variables may update at the same time, and it is not immediately obvious what forecast change was created by what updated variable. However, they can still be useful for thinking about forecast revisions in univariate models. We will therefore start with the simpler univariate case to explain how things work, and then move to the multivariate case afterwards."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note on revisions**: the framework that we are using is designed to decompose changes to forecasts from newly observed datapoints. It can also take into account *revisions* to previously published datapoints, but it does not decompose them separately. Instead, it only shows the aggregate effect of \"revisions\"."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note on `exog` data**: the framework that we are using only decomposes changes to forecasts from newly observed datapoints for *modeled* variables. These are the \"left-hand-side\" variables that in Statsmodels are given in the `endog` arguments. This framework does not decompose or account for changes to unmodeled \"right-hand-side\" variables, like those included in the `exog` argument."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Simple univariate example: AR(1)\n",
    "\n",
    "We will begin with a simple autoregressive model, an AR(1):\n",
    "\n",
    "$$y_t = \\phi y_{t-1} + \\varepsilon_t$$\n",
    "\n",
    "- The parameter $\\phi$ captures the persistence of the series\n",
    "\n",
    "We will use this model to forecast inflation.\n",
    "\n",
    "To make it simpler to describe the forecast updates in this notebook, we will work with inflation data that has been de-meaned, but it is straightforward in practice to augment the model with a mean term.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:07.955618Z",
     "iopub.status.busy": "2022-11-02T17:10:07.955207Z",
     "iopub.status.idle": "2022-11-02T17:10:07.959700Z",
     "shell.execute_reply": "2022-11-02T17:10:07.958665Z"
    }
   },
   "outputs": [],
   "source": [
    "# De-mean the inflation series\n",
    "y = macrodata['infl'] - macrodata['infl'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1: fitting the model on the available dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we'll simulate an out-of-sample exercise, by constructing and fitting our model using all of the data except the last five observations. We'll assume that we haven't observed these values yet, and then in subsequent steps we'll add them back into the analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:07.962988Z",
     "iopub.status.busy": "2022-11-02T17:10:07.962596Z",
     "iopub.status.idle": "2022-11-02T17:10:08.387915Z",
     "shell.execute_reply": "2022-11-02T17:10:08.387249Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pre = y.iloc[:-5]\n",
    "y_pre.plot(figsize=(15, 3), title='Inflation');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To construct forecasts, we first estimate the parameters of the model. This returns a results object that we will be able to use produce forecasts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.393134Z",
     "iopub.status.busy": "2022-11-02T17:10:08.391978Z",
     "iopub.status.idle": "2022-11-02T17:10:08.422642Z",
     "shell.execute_reply": "2022-11-02T17:10:08.421976Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                               SARIMAX Results                                \n",
      "==============================================================================\n",
      "Dep. Variable:                   infl   No. Observations:                  198\n",
      "Model:                 ARIMA(1, 0, 0)   Log Likelihood                -446.407\n",
      "Date:                Wed, 02 Nov 2022   AIC                            896.813\n",
      "Time:                        17:10:08   BIC                            903.390\n",
      "Sample:                    03-31-1959   HQIC                           899.475\n",
      "                         - 06-30-2008                                         \n",
      "Covariance Type:                  opg                                         \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "ar.L1          0.6751      0.043     15.858      0.000       0.592       0.759\n",
      "sigma2         5.3027      0.367     14.459      0.000       4.584       6.022\n",
      "===================================================================================\n",
      "Ljung-Box (L1) (Q):                  15.65   Jarque-Bera (JB):                43.04\n",
      "Prob(Q):                              0.00   Prob(JB):                         0.00\n",
      "Heteroskedasticity (H):               0.85   Skew:                             0.18\n",
      "Prob(H) (two-sided):                  0.50   Kurtosis:                         5.26\n",
      "===================================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n"
     ]
    }
   ],
   "source": [
    "mod_pre = sm.tsa.arima.ARIMA(y_pre, order=(1, 0, 0), trend='n')\n",
    "res_pre = mod_pre.fit()\n",
    "print(res_pre.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating the forecasts from the results object `res` is easy - you can just call the `forecast` method with the number of forecasts you want to construct. In this case, we'll construct four out-of-sample forecasts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.427502Z",
     "iopub.status.busy": "2022-11-02T17:10:08.426370Z",
     "iopub.status.idle": "2022-11-02T17:10:08.735027Z",
     "shell.execute_reply": "2022-11-02T17:10:08.734057Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute the forecasts\n",
    "forecasts_pre = res_pre.forecast(4)\n",
    "\n",
    "# Plot the last 3 years of data and the four out-of-sample forecasts\n",
    "y_pre.iloc[-12:].plot(figsize=(15, 3), label='Data', legend=True)\n",
    "forecasts_pre.plot(label='Forecast', legend=True);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the AR(1) model, it is also easy to manually construct the forecasts. Denoting the last observed variable as $y_T$ and the $h$-step-ahead forecast as $y_{T+h|T}$, we have:\n",
    "\n",
    "$$y_{T+h|T} = \\hat \\phi^h y_T$$\n",
    "\n",
    "Where $\\hat \\phi$ is our estimated value for the AR(1) coefficient. From the summary output above, we can see that this is the first parameter of the model, which we can access from the `params` attribute of the results object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.739026Z",
     "iopub.status.busy": "2022-11-02T17:10:08.738619Z",
     "iopub.status.idle": "2022-11-02T17:10:08.757399Z",
     "shell.execute_reply": "2022-11-02T17:10:08.756549Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        predicted_mean         0\n",
      "2008Q3        3.084388  3.084388\n",
      "2008Q4        2.082323  2.082323\n",
      "2009Q1        1.405812  1.405812\n",
      "2009Q2        0.949088  0.949088\n"
     ]
    }
   ],
   "source": [
    "# Get the estimated AR(1) coefficient\n",
    "phi_hat = res_pre.params[0]\n",
    "\n",
    "# Get the last observed value of the variable\n",
    "y_T = y_pre.iloc[-1]\n",
    "\n",
    "# Directly compute the forecasts at the horizons h=1,2,3,4\n",
    "manual_forecasts = pd.Series([phi_hat * y_T, phi_hat**2 * y_T,\n",
    "                              phi_hat**3 * y_T, phi_hat**4 * y_T],\n",
    "                             index=forecasts_pre.index)\n",
    "\n",
    "# We'll print the two to double-check that they're the same\n",
    "print(pd.concat([forecasts_pre, manual_forecasts], axis=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 2: computing the \"news\" from a new observation\n",
    "\n",
    "Suppose that time has passed, and we have now received another observation. Our dataset is now larger, and we can evaluate our forecast error and produce updated forecasts for the subsequent quarters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.761108Z",
     "iopub.status.busy": "2022-11-02T17:10:08.760752Z",
     "iopub.status.idle": "2022-11-02T17:10:08.773807Z",
     "shell.execute_reply": "2022-11-02T17:10:08.773024Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Forecast error: -10.21\n"
     ]
    }
   ],
   "source": [
    "# Get the next observation after the \"pre\" dataset\n",
    "y_update = y.iloc[-5:-4]\n",
    "\n",
    "# Print the forecast error\n",
    "print('Forecast error: %.2f' % (y_update.iloc[0] - forecasts_pre.iloc[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To compute forecasts based on our updated dataset, we will create an updated results object `res_post` using the `append` method, to append on our new observation to the previous dataset.\n",
    "\n",
    "Note that by default, the `append` method does not re-estimate the parameters of the model. This is exactly what we want here, since we want to isolate the effect on the forecasts of the new information only."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.777793Z",
     "iopub.status.busy": "2022-11-02T17:10:08.777520Z",
     "iopub.status.idle": "2022-11-02T17:10:08.805821Z",
     "shell.execute_reply": "2022-11-02T17:10:08.804862Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2008Q3   -7.121330\n",
      "2008Q4   -4.807732\n",
      "2009Q1   -3.245783\n",
      "2009Q2   -2.191284\n",
      "Freq: Q-DEC, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Create a new results object by passing the new observations to the `append` method\n",
    "res_post = res_pre.append(y_update)\n",
    "\n",
    "# Since we now know the value for 2008Q3, we will only use `res_post` to\n",
    "# produce forecasts for 2008Q4 through 2009Q2\n",
    "forecasts_post = pd.concat([y_update, res_post.forecast('2009Q2')])\n",
    "print(forecasts_post)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case, the forecast error is quite large - inflation was more than 10 percentage points below the AR(1) models' forecast. (This was largely because of large swings in oil prices around the global financial crisis)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To analyse this in more depth, we can use Statsmodels to isolate the effect of the new information - or the \"**news**\" - on our forecasts. This means that we do not yet want to change our model or re-estimate the parameters. Instead, we will use the `news` method that is available in the results objects of state space models.\n",
    "\n",
    "Computing the news in Statsmodels always requires a *previous* results object or dataset, and an *updated* results object or dataset. Here we will use the original results object `res_pre` as the previous results and the `res_post` results object that we just created as the updated results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once we have previous and updated results objects or datasets, we can compute the news by calling the `news` method. Here, we will call `res_pre.news`, and the first argument will be the updated results, `res_post` (however, if you have two results objects, the `news` method could can be called on either one).\n",
    "\n",
    "In addition to specifying the comparison object or dataset as the first argument, there are a variety of other arguments that are accepted. The most important specify the \"impact periods\" that you want to consider. These \"impact periods\" correspond to the forecasted periods of interest; i.e. these dates specify with periods will have forecast revisions decomposed.\n",
    "\n",
    "To specify the impact periods, you must pass two of `start`, `end`, and `periods` (similar to the Pandas `date_range` method). If your time series was a Pandas object with an associated date or period index, then you can pass dates as values for `start` and `end`, as we do below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.809190Z",
     "iopub.status.busy": "2022-11-02T17:10:08.808941Z",
     "iopub.status.idle": "2022-11-02T17:10:08.848807Z",
     "shell.execute_reply": "2022-11-02T17:10:08.848044Z"
    }
   },
   "outputs": [],
   "source": [
    "# Compute the impact of the news on the four periods that we previously\n",
    "# forecasted: 2008Q3 through 2009Q2\n",
    "news = res_pre.news(res_post, start='2008Q3', end='2009Q2')\n",
    "# Note: one alternative way to specify these impact dates is\n",
    "# `start='2008Q3', periods=4`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The variable `news` is an object of the class `NewsResults`, and it contains details about the updates to the data in `res_post` compared to `res_pre`, the new information in the updated dataset, and the impact that the new information had on the forecasts in the period between `start` and `end`.\n",
    "\n",
    "One easy way to summarize the results are with the `summary` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.858964Z",
     "iopub.status.busy": "2022-11-02T17:10:08.858549Z",
     "iopub.status.idle": "2022-11-02T17:10:08.939430Z",
     "shell.execute_reply": "2022-11-02T17:10:08.938564Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                     News                                     \n",
      "==============================================================================\n",
      "Model:                          ARIMA   Original sample:                1959Q1\n",
      "Date:                Wed, 02 Nov 2022                                 - 2008Q2\n",
      "Time:                        17:10:08   Update through:                 2008Q3\n",
      "                                        No. Revisions:                       0\n",
      "                                        No. New datapoints:                  1\n",
      "          Impacts for [impacted variable = infl]         \n",
      "=========================================================\n",
      "impact date estimate (prev) impact of news estimate (new)\n",
      "---------------------------------------------------------\n",
      "     2008Q3            3.08         -10.21          -7.12\n",
      "     2008Q4            2.08          -6.89          -4.81\n",
      "     2009Q1            1.41          -4.65          -3.25\n",
      "     2009Q2            0.95          -3.14          -2.19\n",
      "                  News from updated observations:                  \n",
      "===================================================================\n",
      "update date updated variable   observed forecast (prev)        news\n",
      "-------------------------------------------------------------------\n",
      "     2008Q3             infl      -7.12            3.08      -10.21\n",
      "           Details for [updated variable = infl, impacted variable = infl]           \n",
      "=====================================================================================\n",
      "update date   observed forecast (prev) impact date        news      weight     impact\n",
      "-------------------------------------------------------------------------------------\n",
      "     2008Q3      -7.12            3.08      2008Q3      -10.21         1.0     -10.21\n",
      "                                            2008Q4      -10.21        0.68      -6.89\n",
      "                                            2009Q1      -10.21        0.46      -4.65\n",
      "                                            2009Q2      -10.21        0.31      -3.14\n",
      "=====================================================================================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/hostedtoolcache/Python/3.10.8/x64/lib/python3.10/site-packages/statsmodels/tsa/statespace/news.py:591: FutureWarning: In a future version, `df.iloc[:, i] = newvals` will attempt to set the values inplace instead of always setting a new array. To retain the old behavior, use either `df[df.columns[i]] = newvals` or, if columns are non-unique, `df.isetitem(i, newvals)`\n",
      "  impacts.iloc[:, 0] = impacts.iloc[:, 0].map(str)\n"
     ]
    }
   ],
   "source": [
    "print(news.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Summary output**: the default summary for this news results object printed four tables:\n",
    "\n",
    "1. Summary of the model and datasets\n",
    "2. Details of the news from updated data\n",
    "3. Summary of the impacts of the new information on the forecasts between `start='2008Q3'` and `end='2009Q2'`\n",
    "4. Details of how the updated data led to the impacts on the forecasts between `start='2008Q3'` and `end='2009Q2'`\n",
    "\n",
    "These are described in more detail below.\n",
    "\n",
    "*Notes*:\n",
    "\n",
    "- There are a number of arguments that can be passed to the `summary` method to control this output. Check the documentation / docstring for details.\n",
    "- Table (4), showing details of the updates and impacts, can become quite large if the model is multivariate, there are multiple updates, or a large number of impact dates are selected. It is only shown by default for univariate models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**First table: summary of the model and datasets**\n",
    "\n",
    "The first table, above, shows:\n",
    "\n",
    "- The type of model from which the forecasts were made. Here this is an ARIMA model, since an AR(1) is a special case of an ARIMA(p,d,q) model.\n",
    "- The date and time at which the analysis was computed.\n",
    "- The original sample period, which here corresponds to `y_pre`\n",
    "- The endpoint of the updated sample period, which here is the last date in `y_post`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Second table: the news from updated data**\n",
    "\n",
    "This table simply shows the forecasts from the previous results for observations that were updated in the updated sample.\n",
    "\n",
    "*Notes*:\n",
    "\n",
    "- Our updated dataset `y_post` did not contain any *revisions* to previously observed datapoints. If it had, there would be an additional table showing the previous and updated values of each such revision."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Third table: summary of the impacts of the new information**\n",
    "\n",
    "*Columns*:\n",
    "\n",
    "The third table, above, shows:\n",
    "\n",
    "- The previous forecast for each of the impact dates, in the \"estimate (prev)\" column\n",
    "- The impact that the new information (the \"news\") had on the forecasts for each of the impact dates, in the \"impact of news\" column\n",
    "- The updated forecast for each of the impact dates, in the \"estimate (new)\" column\n",
    "\n",
    "*Notes*:\n",
    "\n",
    "- In multivariate models, this table contains additional columns describing the relevant impacted variable for each row.\n",
    "- Our updated dataset `y_post` did not contain any *revisions* to previously observed datapoints. If it had, there would be additional columns in this table showing the impact of those revisions on the forecasts for the impact dates.\n",
    "- Note that `estimate (new) = estimate (prev) + impact of news`\n",
    "- This table can be accessed independently using the `summary_impacts` method.\n",
    "\n",
    "*In our example*:\n",
    "\n",
    "Notice that in our example, the table shows the values that we computed earlier:\n",
    "\n",
    "- The \"estimate (prev)\" column is identical to the forecasts from our previous model, contained in the `forecasts_pre` variable.\n",
    "- The \"estimate (new)\" column is identical to our `forecasts_post` variable, which contains the observed value for 2008Q3 and the forecasts from the updated model for 2008Q4 - 2009Q2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Fourth table: details of updates and their impacts**\n",
    "\n",
    "The fourth table, above, shows how each new observation translated into specific impacts at each impact date.\n",
    "\n",
    "*Columns*:\n",
    "\n",
    "The first three columns table described the relevant **update** (an \"updated\" is a new observation):\n",
    "\n",
    "- The first column (\"update date\") shows the date of the variable that was updated.\n",
    "- The second column (\"forecast (prev)\") shows the value that would have been forecasted for the update variable at the update date based on the previous results / dataset.\n",
    "- The third column (\"observed\") shows the actual observed value of that updated variable / update date in the updated results / dataset.\n",
    "\n",
    "The last four columns described the **impact** of a given update (an impact is a changed forecast within the \"impact periods\").\n",
    "\n",
    "- The fourth column (\"impact date\") gives the date at which the given update made an impact.\n",
    "- The fifth column (\"news\") shows the \"news\" associated with the given update (this is the same for each impact of a given update, but is just not sparsified by default)\n",
    "- The sixth column (\"weight\") describes the weight that the \"news\" from the given update has on the impacted variable at the impact date. In general, weights will be different between each \"updated variable\" / \"update date\" / \"impacted variable\" / \"impact date\" combination.\n",
    "- The seventh column (\"impact\") shows the impact that the given update had on the given \"impacted variable\" / \"impact date\".\n",
    "\n",
    "*Notes*:\n",
    "\n",
    "- In multivariate models, this table contains additional columns to show the relevant variable that was updated and variable that was impacted for each row. Here, there is only one variable (\"infl\"), so those columns are suppressed to save space.\n",
    "- By default, the updates in this table are \"sparsified\" with blanks, to avoid repeating the same values for \"update date\", \"forecast (prev)\", and \"observed\" for each row of the table. This behavior can be overridden using the `sparsify` argument.\n",
    "- Note that `impact = news * weight`.\n",
    "- This table can be accessed independently using the `summary_details` method.\n",
    "\n",
    "*In our example*:\n",
    "\n",
    "- For the update to 2008Q3 and impact date 2008Q3, the weight is equal to 1. This is because we only have one variable, and once we have incorporated the data for 2008Q3, there is no no remaining ambiguity about the \"forecast\" for this date. Thus all of the \"news\" about this variable at 2008Q3 passes through to the \"forecast\" directly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Addendum: manually computing the news, weights, and impacts\n",
    "\n",
    "For this simple example with a univariate model, it is straightforward to compute all of the values shown above by hand. First, recall the formula for forecasting $y_{T+h|T} = \\phi^h y_T$, and note that it follows that we also have $y_{T+h|T+1} = \\phi^h y_{T+1}$. Finally, note that $y_{T|T+1} = y_T$, because if we know the value of the observations through $T+1$, we know the value of $y_T$.\n",
    "\n",
    "**News**: The \"news\" is nothing more than the forecast error associated with one of the new observations. So the news associated with observation $T+1$ is:\n",
    "\n",
    "$$n_{T+1} = y_{T+1} - y_{T+1|T} = Y_{T+1} - \\phi Y_T$$\n",
    "\n",
    "**Impacts**: The impact of the news is the difference between the updated and previous forecasts, $i_h \\equiv y_{T+h|T+1} - y_{T+h|T}$.\n",
    "\n",
    "- The previous forecasts for $h=1, \\dots, 4$ are: $\\begin{pmatrix} \\phi y_T & \\phi^2 y_T & \\phi^3 y_T & \\phi^4 y_T \\end{pmatrix}'$. \n",
    "- The updated forecasts for $h=1, \\dots, 4$ are: $\\begin{pmatrix} y_{T+1} & \\phi y_{T+1} & \\phi^2 y_{T+1} & \\phi^3 y_{T+1} \\end{pmatrix}'$.\n",
    "\n",
    "The impacts are therefore:\n",
    "\n",
    "$$\\{ i_h \\}_{h=1}^4 = \\begin{pmatrix} y_{T+1} - \\phi y_T \\\\ \\phi (Y_{T+1} - \\phi y_T) \\\\ \\phi^2 (Y_{T+1} - \\phi y_T) \\\\ \\phi^3 (Y_{T+1} - \\phi y_T) \\end{pmatrix}$$\n",
    "\n",
    "**Weights**: To compute the weights, we just need to note that it is immediate that we can rewrite the impacts in terms of the forecast errors, $n_{T+1}$.\n",
    "\n",
    "$$\\{ i_h \\}_{h=1}^4 = \\begin{pmatrix} 1 \\\\ \\phi \\\\ \\phi^2 \\\\ \\phi^3 \\end{pmatrix} n_{T+1}$$\n",
    "\n",
    "The weights are then simply $w = \\begin{pmatrix} 1 \\\\ \\phi \\\\ \\phi^2 \\\\ \\phi^3 \\end{pmatrix}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can check that this is what the `news` method has computed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.943629Z",
     "iopub.status.busy": "2022-11-02T17:10:08.943241Z",
     "iopub.status.idle": "2022-11-02T17:10:08.949869Z",
     "shell.execute_reply": "2022-11-02T17:10:08.949107Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "update date  updated variable\n",
      "2008Q3       infl               -10.205718\n",
      "Name: news, dtype: float64\n",
      "\n",
      "-10.205718\n"
     ]
    }
   ],
   "source": [
    "# Print the news, computed by the `news` method\n",
    "print(news.news)\n",
    "\n",
    "# Manually compute the news\n",
    "print()\n",
    "print((y_update.iloc[0] - phi_hat * y_pre.iloc[-1]).round(6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.953008Z",
     "iopub.status.busy": "2022-11-02T17:10:08.952663Z",
     "iopub.status.idle": "2022-11-02T17:10:08.959711Z",
     "shell.execute_reply": "2022-11-02T17:10:08.958875Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             infl\n",
      "2008Q3 -10.205718\n",
      "2008Q4  -6.890055\n",
      "2009Q1  -4.651595\n",
      "2009Q2  -3.140371\n",
      "\n",
      "2008Q3   -10.205718\n",
      "2008Q4    -6.890055\n",
      "2009Q1    -4.651595\n",
      "2009Q2    -3.140371\n",
      "Freq: Q-DEC, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Print the total impacts, computed by the `news` method\n",
    "# (Note: news.total_impacts = news.revision_impacts + news.update_impacts, but\n",
    "# here there are no data revisions, so total and update impacts are the same)\n",
    "print(news.total_impacts)\n",
    "\n",
    "# Manually compute the impacts\n",
    "print()\n",
    "print(forecasts_post - forecasts_pre)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.962659Z",
     "iopub.status.busy": "2022-11-02T17:10:08.962324Z",
     "iopub.status.idle": "2022-11-02T17:10:08.976609Z",
     "shell.execute_reply": "2022-11-02T17:10:08.975891Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "impact date                  2008Q3    2008Q4    2009Q1    2009Q2\n",
      "impacted variable              infl      infl      infl      infl\n",
      "update date updated variable                                     \n",
      "2008Q3      infl                1.0  0.675117  0.455783  0.307707"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "[1.       0.675117 0.455783 0.307707]\n"
     ]
    }
   ],
   "source": [
    "# Print the weights, computed by the `news` method\n",
    "print(news.weights)\n",
    "\n",
    "# Manually compute the weights\n",
    "print()\n",
    "print(np.array([1, phi_hat, phi_hat**2, phi_hat**3]).round(6))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Multivariate example: dynamic factor\n",
    "\n",
    "In this example, we'll consider forecasting monthly core price inflation based on the Personal Consumption Expenditures (PCE) price index and the Consumer Price Index (CPI), using a Dynamic Factor model. Both of these measures track prices in the US economy and are based on similar source data, but they have a number of definitional differences. Nonetheless, they track each other relatively well, so modeling them jointly using a single dynamic factor seems reasonable.\n",
    "\n",
    "One reason that this kind of approach can be useful is that the CPI is released earlier in the month than the PCE. One the CPI is released, therefore, we can update our dynamic factor model with that additional datapoint, and obtain an improved forecast for that month's PCE release. A more involved version of this kind of analysis is available in Knotek and Zaman (2017)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We start by downloading the core CPI and PCE price index data from [FRED](https://fred.stlouisfed.org/), converting them to annualized monthly inflation rates, removing two outliers, and de-meaning each series (the dynamic factor model does not "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:08.980056Z",
     "iopub.status.busy": "2022-11-02T17:10:08.979719Z",
     "iopub.status.idle": "2022-11-02T17:10:09.243888Z",
     "shell.execute_reply": "2022-11-02T17:10:09.243056Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas_datareader as pdr\n",
    "levels = pdr.get_data_fred(['PCEPILFE', 'CPILFESL'], start='1999', end='2019').to_period('M')\n",
    "infl = np.log(levels).diff().iloc[1:] * 1200\n",
    "infl.columns = ['PCE', 'CPI']\n",
    "\n",
    "# Remove two outliers and de-mean the series\n",
    "infl['PCE'].loc['2001-09':'2001-10'] = np.nan"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To show how this works, we'll imagine that it is April 14, 2017, which is the data of the March 2017 CPI release. So that we can show the effect of multiple updates at once, we'll assume that we haven't updated our data since the end of January, so that:\n",
    "\n",
    "- Our **previous dataset** will consist of all values for the PCE and CPI through January 2017\n",
    "- Our **updated dataset** will additionally incorporate the CPI for February and March 2017 and the PCE data for February 2017. But it will not yet the PCE (the March 2017 PCE price index was not released until May 1, 2017)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:09.248248Z",
     "iopub.status.busy": "2022-11-02T17:10:09.247986Z",
     "iopub.status.idle": "2022-11-02T17:10:09.262736Z",
     "shell.execute_reply": "2022-11-02T17:10:09.261960Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              PCE       CPI\n",
      "DATE                       \n",
      "2016-09  1.385688  2.022262\n",
      "2016-10  1.777645  1.445830\n",
      "2016-11  0.584472  1.631694\n",
      "2016-12  1.572165  2.109728\n",
      "2017-01  3.093392  2.623570\n"
     ]
    }
   ],
   "source": [
    "# Previous dataset runs through 2017-02\n",
    "y_pre = infl.loc[:'2017-01'].copy()\n",
    "const_pre = np.ones(len(y_pre))\n",
    "print(y_pre.tail())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:09.266836Z",
     "iopub.status.busy": "2022-11-02T17:10:09.266489Z",
     "iopub.status.idle": "2022-11-02T17:10:09.274300Z",
     "shell.execute_reply": "2022-11-02T17:10:09.273484Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              PCE       CPI\n",
      "DATE                       \n",
      "2016-11  0.584472  1.631694\n",
      "2016-12  1.572165  2.109728\n",
      "2017-01  3.093392  2.623570\n",
      "2017-02  2.337166  2.541355\n",
      "2017-03       NaN -0.258197\n"
     ]
    }
   ],
   "source": [
    "# For the updated dataset, we'll just add in the\n",
    "# CPI value for 2017-03\n",
    "y_post = infl.loc[:'2017-03'].copy()\n",
    "y_post.loc['2017-03', 'PCE'] = np.nan\n",
    "const_post = np.ones(len(y_post))\n",
    "\n",
    "# Notice the missing value for PCE in 2017-03\n",
    "print(y_post.tail())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We chose this particular example because in March 2017, core CPI prices fell for the first time since 2010, and this information may be useful in forecast core PCE prices for that month. The graph below shows the CPI and PCE price data as it would have been observed on April 14th$^\\dagger$.\n",
    "\n",
    "-----\n",
    "\n",
    "$\\dagger$ This statement is not entirely true, because both the CPI and PCE price indexes can be revised to a certain extent after the fact. As a result, the series that we're pulling are not exactly like those observed on April 14, 2017. This could be fixed by pulling the archived data from [ALFRED](https://alfred.stlouisfed.org/) instead of [FRED](https://fred.stlouisfed.org/), but the data we have is good enough for this tutorial."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:09.277659Z",
     "iopub.status.busy": "2022-11-02T17:10:09.277307Z",
     "iopub.status.idle": "2022-11-02T17:10:09.623560Z",
     "shell.execute_reply": "2022-11-02T17:10:09.622608Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the updated dataset\n",
    "fig, ax = plt.subplots(figsize=(15, 3))\n",
    "y_post.plot(ax=ax)\n",
    "ax.hlines(0, '2009', '2017-06', linewidth=1.0)\n",
    "ax.set_xlim('2009', '2017-06');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To perform the exercise, we first construct and fit a `DynamicFactor` model. Specifically:\n",
    "\n",
    "- We are using a single dynamic factor (`k_factors=1`)\n",
    "- We are modeling the factor's dynamics with an AR(6) model (`factor_order=6`)\n",
    "- We have included a vector of ones as an exogenous variable (`exog=const_pre`), because the inflation series we are working with are not mean-zero."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:09.627541Z",
     "iopub.status.busy": "2022-11-02T17:10:09.627046Z",
     "iopub.status.idle": "2022-11-02T17:10:11.471713Z",
     "shell.execute_reply": "2022-11-02T17:10:11.470947Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RUNNING THE L-BFGS-B CODE\n",
      "\n",
      "           * * *\n",
      "\n",
      "Machine precision = 2.220D-16\n",
      " N =           12     M =           10\n",
      "\n",
      "At X0         0 variables are exactly at the bounds\n",
      "\n",
      "At iterate    0    f=  4.82073D+00    |proj g|=  3.19124D-01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " This problem is unconstrained.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "At iterate    5    f=  4.63641D+00    |proj g|=  3.14144D-01\n",
      "  ys=-8.605E-01  -gs= 5.450E-01 BFGS update SKIPPED\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "At iterate   10    f=  2.95557D+00    |proj g|=  2.77682D-01\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "At iterate   15    f=  2.57945D+00    |proj g|=  2.07981D-01\n",
      "\n",
      "At iterate   20    f=  2.43842D+00    |proj g|=  8.68367D-02\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "At iterate   25    f=  2.43091D+00    |proj g|=  1.15711D-02\n",
      "\n",
      "At iterate   30    f=  2.42946D+00    |proj g|=  8.06022D-03\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "At iterate   35    f=  2.42937D+00    |proj g|=  8.24406D-05\n",
      "\n",
      "           * * *\n",
      "\n",
      "Tit   = total number of iterations\n",
      "Tnf   = total number of function evaluations\n",
      "Tnint = total number of segments explored during Cauchy searches\n",
      "Skip  = number of BFGS updates skipped\n",
      "Nact  = number of active bounds at final generalized Cauchy point\n",
      "Projg = norm of the final projected gradient\n",
      "F     = final function value\n",
      "\n",
      "           * * *\n",
      "\n",
      "   N    Tit     Tnf  Tnint  Skip  Nact     Projg        F\n",
      "   12     37     51      1     1     0   2.265D-05   2.429D+00\n",
      "  F =   2.4293664173073029     \n",
      "\n",
      "CONVERGENCE: REL_REDUCTION_OF_F_<=_FACTR*EPSMCH             \n",
      "                                   Statespace Model Results                                  \n",
      "=============================================================================================\n",
      "Dep. Variable:                        ['PCE', 'CPI']   No. Observations:                  216\n",
      "Model:             DynamicFactor(factors=1, order=6)   Log Likelihood                -524.743\n",
      "                                      + 1 regressors   AIC                           1073.486\n",
      "Date:                               Wed, 02 Nov 2022   BIC                           1113.990\n",
      "Time:                                       17:10:11   HQIC                          1089.850\n",
      "Sample:                                   02-28-1999                                         \n",
      "                                        - 01-31-2017                                         \n",
      "Covariance Type:                                 opg                                         \n",
      "===================================================================================\n",
      "Ljung-Box (L1) (Q):             4.33, 0.55   Jarque-Bera (JB):         12.52, 10.68\n",
      "Prob(Q):                        0.04, 0.46   Prob(JB):                   0.00, 0.00\n",
      "Heteroskedasticity (H):         0.57, 0.47   Skew:                      0.16, -0.14\n",
      "Prob(H) (two-sided):            0.02, 0.00   Kurtosis:                   4.13, 4.05\n",
      "                           Results for equation PCE                           \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "loading.f1     0.5407      0.061      8.836      0.000       0.421       0.661\n",
      "beta.const     1.7154      0.094     18.269      0.000       1.531       1.899\n",
      "                           Results for equation CPI                           \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "loading.f1     0.9033      0.104      8.717      0.000       0.700       1.106\n",
      "beta.const     1.9620      0.136     14.375      0.000       1.694       2.230\n",
      "                        Results for factor equation f1                        \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "L1.f1          0.1246      0.069      1.805      0.071      -0.011       0.260\n",
      "L2.f1          0.1823      0.072      2.543      0.011       0.042       0.323\n",
      "L3.f1          0.0177      0.073      0.244      0.807      -0.125       0.160\n",
      "L4.f1         -0.0700      0.078     -0.893      0.372      -0.224       0.084\n",
      "L5.f1          0.1561      0.068      2.304      0.021       0.023       0.289\n",
      "L6.f1          0.1376      0.075      1.838      0.066      -0.009       0.284\n",
      "                           Error covariance matrix                            \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "sigma2.PCE     0.5517      0.066      8.381      0.000       0.423       0.681\n",
      "sigma2.CPI  3.132e-10      0.149   2.11e-09      1.000      -0.291       0.291\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n"
     ]
    }
   ],
   "source": [
    "mod_pre = sm.tsa.DynamicFactor(y_pre, exog=const_pre, k_factors=1, factor_order=6)\n",
    "res_pre = mod_pre.fit()\n",
    "print(res_pre.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With the fitted model in hand, we now construct the news and impacts associated with observing the CPI for March 2017. The updated data is for February 2017 and part of March 2017, and we'll examining the impacts on both March and April.\n",
    "\n",
    "In the univariate example, we first created an updated results object, and then passed that to the `news` method. Here, we're creating the news by directly passing the updated dataset.\n",
    "\n",
    "Notice that:\n",
    "\n",
    "1. `y_post` contains the entire updated dataset (not just the new datapoints)\n",
    "2. We also had to pass an updated `exog` array. This array must cover **both**:\n",
    "    - The entire period associated with `y_post`\n",
    "    - Any additional datapoints after the end of `y_post` through the last impact date, specified by `end`\n",
    "\n",
    "   Here, `y_post` ends in March 2017, so we needed our `exog` to extend one more period, to April 2017."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:11.476921Z",
     "iopub.status.busy": "2022-11-02T17:10:11.475773Z",
     "iopub.status.idle": "2022-11-02T17:10:11.553757Z",
     "shell.execute_reply": "2022-11-02T17:10:11.552968Z"
    }
   },
   "outputs": [],
   "source": [
    "# Create the news results\n",
    "# Note\n",
    "const_post_plus1 = np.ones(len(y_post) + 1)\n",
    "news = res_pre.news(y_post, exog=const_post_plus1, start='2017-03', end='2017-04')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> **Note**:\n",
    ">\n",
    "> In the univariate example, above, we first constructed a new results object, and then passed that to the `news` method. We could have done that here too, although there is an extra step required. Since we are requesting an impact for a period beyond the end of `y_post`, we would still need to pass the additional value for the `exog` variable during that period to `news`:\n",
    "> \n",
    "> ```python\n",
    "res_post = res_pre.apply(y_post, exog=const_post)\n",
    "news = res_pre.news(res_post, exog=[1.], start='2017-03', end='2017-04')\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have computed the `news`, printing `summary` is a convenient way to see the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:11.559393Z",
     "iopub.status.busy": "2022-11-02T17:10:11.558166Z",
     "iopub.status.idle": "2022-11-02T17:10:11.687306Z",
     "shell.execute_reply": "2022-11-02T17:10:11.686625Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                     News                                     \n",
      "==============================================================================\n",
      "Model:                  DynamicFactor   Original sample:               1999-02\n",
      "Date:                Wed, 02 Nov 2022                                - 2017-01\n",
      "Time:                        17:10:11   Update through:                2017-04\n",
      "                                        No. Revisions:                       0\n",
      "                                        No. New datapoints:                  3\n",
      "                                  Impacts                                  \n",
      "===========================================================================\n",
      "impact date impacted variable estimate (prev) impact of news estimate (new)\n",
      "---------------------------------------------------------------------------\n",
      "    2017-03               CPI            2.07          -2.33          -0.26\n",
      "                          PCE            1.78          -1.39           0.39\n",
      "    2017-04               CPI            1.90          -0.23           1.67\n",
      "                          PCE            1.68          -0.14           1.54\n",
      "                  News from updated observations:                  \n",
      "===================================================================\n",
      "update date updated variable   observed forecast (prev)        news\n",
      "-------------------------------------------------------------------\n",
      "    2017-02              CPI       2.54            2.24        0.30\n",
      "                         PCE       2.34            1.88        0.46\n",
      "    2017-03              CPI      -0.26            2.07       -2.33\n",
      "===================================================================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/hostedtoolcache/Python/3.10.8/x64/lib/python3.10/site-packages/statsmodels/tsa/statespace/news.py:594: FutureWarning: In a future version, `df.iloc[:, i] = newvals` will attempt to set the values inplace instead of always setting a new array. To retain the old behavior, use either `df[df.columns[i]] = newvals` or, if columns are non-unique, `df.isetitem(i, newvals)`\n",
      "  impacts.iloc[:, :2] = impacts.iloc[:, :2].applymap(str)\n"
     ]
    }
   ],
   "source": [
    "# Show the summary of the news results\n",
    "print(news.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because we have multiple variables, by default the summary only shows the news from updated data along and the total impacts.\n",
    "\n",
    "From the first table, we can see that our updated dataset contains three new data points, with most of the \"news\" from these data coming from the very low reading in March 2017.\n",
    "\n",
    "The second table shows that these three datapoints substantially impacted the estimate for PCE in March 2017 (which was not yet observed). This estimate revised down by nearly 1.5 percentage points.\n",
    "\n",
    "The updated data also impacted the forecasts in the first out-of-sample month, April 2017. After incorporating the new data, the model's forecasts for CPI and PCE inflation in that month revised down 0.29 and 0.17 percentage point, respectively."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While these tables show the \"news\" and the total impacts, they do not show how much of each impact was caused by each updated datapoint. To see that information, we need to look at the details tables.\n",
    "\n",
    "One way to see the details tables is to pass `include_details=True` to the `summary` method. To avoid repeating the tables above, however, we'll just call the `summary_details` method directly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-11-02T17:10:11.692369Z",
     "iopub.status.busy": "2022-11-02T17:10:11.691223Z",
     "iopub.status.idle": "2022-11-02T17:10:11.736220Z",
     "shell.execute_reply": "2022-11-02T17:10:11.735604Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 Details for [updated variable = CPI]                                 \n",
      "======================================================================================================\n",
      "update date   observed forecast (prev) impact date impacted variable        news     weight     impact\n",
      "------------------------------------------------------------------------------------------------------\n",
      "    2017-02       2.54            2.24     2017-04               CPI        0.30       0.18       0.06\n",
      "                                                                 PCE        0.30       0.11       0.03\n",
      "    2017-03      -0.26            2.07     2017-03               CPI       -2.33       1.00      -2.33\n",
      "                                                                 PCE       -2.33       0.60      -1.39\n",
      "                                           2017-04               CPI       -2.33       0.12      -0.29\n",
      "                                                                 PCE       -2.33       0.07      -0.17\n",
      "======================================================================================================\n"
     ]
    }
   ],
   "source": [
    "print(news.summary_details())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This table shows that most of the revisions to the estimate of PCE in April 2017, described above, came from the news associated with the CPI release in March 2017. By contrast, the CPI release in February had only a little effect on the April forecast, and the PCE release in February had essentially no effect."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bibliography\n",
    "\n",
    "Bańbura, Marta, Domenico Giannone, and Lucrezia Reichlin. \"Nowcasting.\" The Oxford Handbook of Economic Forecasting. July 8, 2011.\n",
    "\n",
    "Bańbura, Marta, Domenico Giannone, Michele Modugno, and Lucrezia Reichlin. \"Now-casting and the real-time data flow.\" In Handbook of economic forecasting, vol. 2, pp. 195-237. Elsevier, 2013.\n",
    "\n",
    "Bańbura, Marta, and Michele Modugno. \"Maximum likelihood estimation of factor models on datasets with arbitrary pattern of missing data.\" Journal of Applied Econometrics 29, no. 1 (2014): 133-160.\n",
    "\n",
    "Knotek, Edward S., and Saeed Zaman. \"Nowcasting US headline and core inflation.\" Journal of Money, Credit and Banking 49, no. 5 (2017): 931-968."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}