diff --git a/.gitignore b/.gitignore index c2b68b7b..bfd1f8af 100644 --- a/.gitignore +++ b/.gitignore @@ -10,3 +10,4 @@ !spm_threshold_agi.csv **/_build !population_by_state.csv +.ipynb_checkpoints \ No newline at end of file diff --git a/changelog_entry.yaml b/changelog_entry.yaml index e69de29b..f73e252f 100644 --- a/changelog_entry.yaml +++ b/changelog_entry.yaml @@ -0,0 +1,4 @@ +- bump: minor + changes: + changed: + - Worked on SCF imputation. diff --git a/docs/scf/SCF_imputation.ipynb b/docs/scf/SCF_imputation.ipynb new file mode 100644 index 00000000..1d4fcd47 --- /dev/null +++ b/docs/scf/SCF_imputation.ipynb @@ -0,0 +1,2652 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 8, + "id": "8abd5097", + "metadata": {}, + "outputs": [], + "source": [ + "#!pip install git+https://github.com/PSLmodels/scf.git\n", + "#!pip install git+https://github.com/pslmodels/scf/\n", + "#!pip install quantile_forest" + ] + }, + { + "cell_type": "markdown", + "id": "856eb849-74f8-42e5-922f-546fc7c5d21f", + "metadata": {}, + "source": [ + "## Imputation of the Survey of Consumer Finances with Quantile Regression Forests" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "id": "c8360b7d-0da3-44cd-9aa1-f55a69c9032d", + "metadata": {}, + "outputs": [], + "source": [ + "import microdf as mdf\n", + "import scf\n", + "import pandas as pd\n", + "from typing import Union\n", + "from sklearn.model_selection import train_test_split\n", + "from sklearn.model_selection import KFold\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "import seaborn as sns\n", + "\n", + "VALID_YEARS = [\n", + " 1989,\n", + " 1992,\n", + " 1995,\n", + " 1998,\n", + " 2001,\n", + " 2004,\n", + " 2007,\n", + " 2010,\n", + " 2013,\n", + " 2016,\n", + " 2019,\n", + "]\n", + "\n", + "def scf_url(year: int) -> str:\n", + " \"\"\" Returns the URL of the SCF summary microdata zip file for a year.\n", + "\n", + " :param year: Year of SCF summary microdata to retrieve.\n", + " :type year: int\n", + " :return: URL of summary microdata zip file for the given year.\n", + " :rtype: str\n", + " \"\"\"\n", + " assert year in VALID_YEARS, \"The SCF is not available for \" + str(year)\n", + " return (\n", + " \"https://www.federalreserve.gov/econres/files/scfp\"\n", + " + str(year)\n", + " + \"s.zip\"\n", + " )\n", + "\n", + "\n", + "def load_single_scf(year: int, columns: list) -> pd.DataFrame:\n", + " \"\"\" Loads SCF summary microdata for a given year and set of columns.\n", + "\n", + " :param year: Year of SCF summary microdata to retrieve.\n", + " :type year: int\n", + " :param columns: List of columns. The weight column `wgt` is always\n", + " returned. Defaults to all columns in the summary dataset.\n", + " :type columns: list\n", + " :return: SCF summary microdata for the given year.\n", + " :rtype: pd.DataFrame\n", + " \"\"\"\n", + " # Add wgt to all returns.\n", + " if columns is not None:\n", + " columns = list(set(columns) | set([\"wgt\"]))\n", + " return mdf.read_stata_zip(scf_url(year), columns=columns)\n", + "\n", + "\n", + "def load(\n", + " years: list = VALID_YEARS,\n", + " columns: list = None,\n", + " as_microdataframe: bool = False,\n", + ") -> Union[pd.DataFrame, mdf.MicroDataFrame]:\n", + " \"\"\" Loads SCF summary microdata for a set of years and columns.\n", + "\n", + " :param years: Year(s) to load SCF data for. Can be a list or single number.\n", + " Defaults to all available years, starting with 1989.\n", + " :type years: list\n", + " :param columns: List of columns. The weight column `wgt` is always returned.\n", + " :type columns: list\n", + " :param as_microdataframe: Whether to return as a MicroDataFrame with\n", + " weight set, defaults to False.\n", + " :type as_microdataframe: bool\n", + " :return: SCF summary microdata for the set of years.\n", + " :rtype: Union[pd.DataFrame, mdf.MicroDataFrame]\n", + " \"\"\"\n", + " # Make cols a list if a single column is passed.\n", + " if columns is not None:\n", + " columns = mdf.listify(columns)\n", + " # If years is a single year rather than a list, don't use a loop.\n", + " if isinstance(years, int):\n", + " res = load_single_scf(years, columns)\n", + " # Otherwise append to a list within a loop, and concatenate.\n", + " else:\n", + " scfs = []\n", + " for year in years:\n", + " tmp = load_single_scf(year, columns)\n", + " tmp[\"year\"] = year\n", + " scfs.append(tmp)\n", + " res = pd.concat(scfs)\n", + " # Return as a MicroDataFrame or DataFrame.\n", + " if as_microdataframe:\n", + " return mdf.MicroDataFrame(res, weights=\"wgt\")\n", + " return res" + ] + }, + { + "cell_type": "markdown", + "id": "c108b5d3", + "metadata": {}, + "source": [ + "## CPS Columns of Interest" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "id": "5f38cba4", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Files in ZIP archive: ['hhpub20.csv', 'ffpub20.csv', 'pppub20.csv', 'asec_csv_repwgt_2020.csv']\n", + "Person Data Columns: Index(['PERIDNUM', 'PH_SEQ', 'P_SEQ', 'A_LINENO', 'PF_SEQ', 'PHF_SEQ',\n", + " 'OED_TYP1', 'OED_TYP2', 'OED_TYP3', 'PERRP',\n", + " ...\n", + " 'M5G_CBST', 'M5G_DSCP', 'CLWK', 'DEP_STAT', 'FILEDATE', 'FILESTAT',\n", + " 'LJCW', 'NOEMP', 'WECLW', 'YYYYMM'],\n", + " dtype='object', length=840)\n", + "Family Data Columns: Index(['FPOVCUT', 'FPERSONS', 'FHEADIDX', 'FSPOUIDX', 'FOWNU6', 'FRELU6',\n", + " 'FKIND', 'FKINDEX', 'FTYPE', 'FRELU18', 'FOWNU18', 'FLASTIDX',\n", + " 'FMLASIDX', 'FH_SEQ', 'FAMLIS', 'FANNVAL', 'FCSPVAL', 'FDISVAL',\n", + " 'FDIVVAL', 'FDSTVAL', 'FEARNVAL', 'FEDVAL', 'FFINVAL', 'FFPOS',\n", + " 'FFRVAL', 'FHIP_VAL', 'FHIP_VAL2', 'FINC_ANN', 'FINC_CSP', 'FINC_DIS',\n", + " 'FINC_DIV', 'FINC_DST', 'FINC_ED', 'FINC_FIN', 'FINC_FR', 'FINC_INT',\n", + " 'FINC_OI', 'FINC_PAW', 'FINC_PEN', 'FINC_RNT', 'FINC_SE', 'FINC_SS',\n", + " 'FINC_SSI', 'FINC_SUR', 'FINC_UC', 'FINC_VET', 'FINC_WC', 'FINC_WS',\n", + " 'FINTVAL', 'FMED_VAL', 'FMOOP', 'FMOOP2', 'FOIVAL', 'FOTC_VAL',\n", + " 'FOTHVAL', 'FPAWVAL', 'FPCTCUT', 'FPENVAL', 'FRECORD', 'FRNTVAL',\n", + " 'FRSPOV', 'FRSPPCT', 'FSEVAL', 'FSPANISH', 'FSSIVAL', 'FSSVAL',\n", + " 'FSUP_WGT', 'FSURVAL', 'FTOTVAL', 'FTOT_R', 'FUCVAL', 'FVETVAL',\n", + " 'FWCVAL', 'FWSVAL', 'F_MV_FS', 'F_MV_SL', 'I_FHIPVAL', 'I_FHIPVAL2',\n", + " 'I_FMEDVAL', 'I_FMOOP', 'I_FMOOP2', 'I_FOTCVAL', 'POVLL', 'FILEDATE',\n", + " 'YYYYMM'],\n", + " dtype='object')\n", + "Household Data Columns: Index(['H_IDNUM', 'GEREG', 'GESTFIPS', 'GEDIV', 'HRHTYPE', 'HEFAMINC',\n", + " 'H_MONTH', 'H_YEAR', 'H_TENURE', 'H_HHNUM',\n", + " ...\n", + " 'THPROP_VAL', 'GTCBSA', 'GTCO', 'GTCBSAST', 'GTCBSASZ', 'GTCSA',\n", + " 'GTMETSTA', 'GTINDVPC', 'FILEDATE', 'YYYYMM'],\n", + " dtype='object', length=134)\n" + ] + } + ], + "source": [ + "import requests\n", + "from io import BytesIO\n", + "from zipfile import ZipFile\n", + "import pandas as pd\n", + "\n", + "# URL for the 2019 CPS dataset\n", + "CPS_2019_URL = \"https://www2.census.gov/programs-surveys/cps/datasets/2020/march/asecpub20csv.zip\"\n", + "\n", + "# Download the zip file\n", + "response = requests.get(CPS_2019_URL, stream=True)\n", + "if response.status_code == 200:\n", + " with ZipFile(BytesIO(response.content)) as zipfile:\n", + " # List all files in the ZIP archive\n", + " file_list = zipfile.namelist()\n", + " print(\"Files in ZIP archive:\", file_list)\n", + "\n", + " # Load each dataset\n", + " with zipfile.open(\"pppub20.csv\") as f:\n", + " person_df = pd.read_csv(f)\n", + " with zipfile.open(\"ffpub20.csv\") as f:\n", + " family_df = pd.read_csv(f)\n", + " with zipfile.open(\"hhpub20.csv\") as f:\n", + " household_df = pd.read_csv(f)\n", + "\n", + " # Display the first few rows of each dataset\n", + " print(\"Person Data Columns:\", person_df.columns)\n", + " print(\"Family Data Columns:\", family_df.columns)\n", + " print(\"Household Data Columns:\", household_df.columns)\n", + "else:\n", + " print(\"Failed to download the dataset.\")\n" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "id": "4ddd923e", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "PERIDNUM\n", + "H_IDNUM\n" + ] + } + ], + "source": [ + "for column in person_df.columns:\n", + " print(column)\n", + " break # remove to see all columns in dataset\n", + "\n", + "for column in household_df.columns:\n", + " print(column)\n", + " break # remove to see all columns in dataset\n", + "\n", + "# see https://www2.census.gov/programs-surveys/cps/methodology/PublicUseDocumentation_final.pdf and\n", + "# https://www2.census.gov/programs-surveys/cps/datasets/2020/march/ASEC2020ddl_pub_full.pdf " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8079ad3c", + "metadata": {}, + "outputs": [], + "source": [ + "# Messy, just some CPS variables for reference\n", + "\n", + "cps_predictors = [\"is_household_head\",\n", + " \"age\",\n", + " \"is_male\",\n", + " \"tenure_type\",\n", + " \"employment_income\",\n", + " \"self_employment_income\",\n", + " \"social_security\",\n", + " \"pension_income\",\n", + " \"household_size\",]\n", + "\n", + "\n", + "cps[\"own_children_in_household\"] = tmp.children.fillna(0)\n", + "\n", + "cps[\"has_marketplace_health_coverage\"] = person.MRK == 1\n", + "\n", + "cps[\"cps_race\"] = person.PRDTRACE\n", + "cps[\"is_hispanic\"] = person.PRDTHSP != 0\n", + "\n", + "cps[\"is_widowed\"] = person.A_MARITL == 4\n", + "cps[\"is_separated\"] = person.A_MARITL == 6\n", + "# High school or college/university enrollment status.\n", + "cps[\"is_full_time_college_student\"] = person.A_HSCOL == 2\n", + "\n", + "# Assign CPS variables.\n", + "cps[\"employment_income\"] = person.WSAL_VAL\n", + "\n", + "cps[\"weekly_hours_worked\"] = person.HRSWK * person.WKSWORK / 52\n", + "\n", + "cps[\"taxable_interest_income\"] = person.INT_VAL * (\n", + " p[\"taxable_interest_fraction\"]\n", + ")\n", + "cps[\"tax_exempt_interest_income\"] = person.INT_VAL * (\n", + " 1 - p[\"taxable_interest_fraction\"]\n", + ")\n", + "cps[\"self_employment_income\"] = person.SEMP_VAL\n", + "cps[\"farm_income\"] = person.FRSE_VAL\n", + "cps[\"qualified_dividend_income\"] = person.DIV_VAL * (\n", + " p[\"qualified_dividend_fraction\"]\n", + ")\n", + "cps[\"non_qualified_dividend_income\"] = person.DIV_VAL * (\n", + " 1 - p[\"qualified_dividend_fraction\"]\n", + ")\n", + "cps[\"rental_income\"] = person.RNT_VAL\n", + "# Assign Social Security retirement benefits if at least 62.\n", + "MINIMUM_RETIREMENT_AGE = 62\n", + "cps[\"social_security_retirement\"] = np.where(\n", + " person.A_AGE >= MINIMUM_RETIREMENT_AGE, person.SS_VAL, 0\n", + ")\n", + "# Otherwise assign them to Social Security disability benefits.\n", + "cps[\"social_security_disability\"] = (\n", + " person.SS_VAL - cps[\"social_security_retirement\"]\n", + ")\n", + "# Provide placeholders for other Social Security inputs to avoid creating\n", + "# NaNs as they're uprated.\n", + "cps[\"social_security_dependents\"] = np.zeros_like(\n", + " cps[\"social_security_retirement\"]\n", + ")\n", + "cps[\"social_security_survivors\"] = np.zeros_like(\n", + " cps[\"social_security_retirement\"]\n", + ")\n", + "cps[\"unemployment_compensation\"] = person.UC_VAL" + ] + }, + { + "cell_type": "markdown", + "id": "dc6c9274-694d-4208-9da9-508385974942", + "metadata": {}, + "source": [ + "## SCF Data Load and Preprocessing" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "id": "97611bb7-3aa5-40fb-9c20-8f2b41b4e496", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
yy1y1wgthhsexageagecleducedclmarriedkids...inccatassetcatninccatninc2catnwpctlecatincpctlecatnincpctlecatincqrtcatnincqrtcatyear
257765188518824893.41669124538220...3131255222019
13143264426444439.239368150312411...6663111011442019
2647653285328238.184735176612410...6663121212442019
2842757215721364.238949176613420...6663121212442019
237194772477255960.043856155412411...4342677332019
\n", + "

5 rows × 358 columns

\n", + "
" + ], + "text/plain": [ + " yy1 y1 wgt hhsex age agecl educ edcl married kids \\\n", + "25776 5188 51882 4893.416691 2 45 3 8 2 2 0 \n", + "13143 2644 26444 439.239368 1 50 3 12 4 1 1 \n", + "26476 5328 53282 38.184735 1 76 6 12 4 1 0 \n", + "28427 5721 57213 64.238949 1 76 6 13 4 2 0 \n", + "23719 4772 47725 5960.043856 1 55 4 12 4 1 1 \n", + "\n", + " ... inccat assetcat ninccat ninc2cat nwpctlecat incpctlecat \\\n", + "25776 ... 3 1 3 1 2 5 \n", + "13143 ... 6 6 6 3 11 10 \n", + "26476 ... 6 6 6 3 12 12 \n", + "28427 ... 6 6 6 3 12 12 \n", + "23719 ... 4 3 4 2 6 7 \n", + "\n", + " nincpctlecat incqrtcat nincqrtcat year \n", + "25776 5 2 2 2019 \n", + "13143 11 4 4 2019 \n", + "26476 12 4 4 2019 \n", + "28427 12 4 4 2019 \n", + "23719 7 3 3 2019 \n", + "\n", + "[5 rows x 358 columns]" + ] + }, + "execution_count": 13, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# focus on 2019 SCF data\n", + "data = load([VALID_YEARS[-1]])\n", + "\n", + "# split data into train and test sets\n", + "train_df, test_df = train_test_split(data, test_size=0.2, train_size=0.8)\n", + "\n", + "train_df.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "id": "5e9fef5f-9454-4809-b961-db1340575406", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "yy1\n" + ] + } + ], + "source": [ + "for column in train_df.columns:\n", + " print(column)\n", + " break # remove to see all columns in dataset" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "id": "9977b76c-240a-42b0-b775-a532244448cb", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
incomeagemarriedkidseducracelf
count2.310800e+0423108.00000023108.00000023108.00000023108.00000023108.00000023108.000000
mean1.171661e+0653.2386621.3767090.75125510.2520341.5486410.751471
std1.292835e+0716.2253800.4845721.1274482.7080351.0568370.432169
min0.000000e+0018.0000001.0000000.000000-1.0000001.0000000.000000
25%4.248891e+0440.7500001.0000000.0000008.0000001.0000001.000000
50%9.264943e+0454.0000001.0000000.00000011.0000001.0000001.000000
75%2.325088e+0565.0000002.0000001.00000012.0000002.0000001.000000
max8.156336e+0895.0000002.0000007.00000014.0000005.0000001.000000
\n", + "
" + ], + "text/plain": [ + " income age married kids educ \\\n", + "count 2.310800e+04 23108.000000 23108.000000 23108.000000 23108.000000 \n", + "mean 1.171661e+06 53.238662 1.376709 0.751255 10.252034 \n", + "std 1.292835e+07 16.225380 0.484572 1.127448 2.708035 \n", + "min 0.000000e+00 18.000000 1.000000 0.000000 -1.000000 \n", + "25% 4.248891e+04 40.750000 1.000000 0.000000 8.000000 \n", + "50% 9.264943e+04 54.000000 1.000000 0.000000 11.000000 \n", + "75% 2.325088e+05 65.000000 2.000000 1.000000 12.000000 \n", + "max 8.156336e+08 95.000000 2.000000 7.000000 14.000000 \n", + "\n", + " race lf \n", + "count 23108.000000 23108.000000 \n", + "mean 1.548641 0.751471 \n", + "std 1.056837 0.432169 \n", + "min 1.000000 0.000000 \n", + "25% 1.000000 1.000000 \n", + "50% 1.000000 1.000000 \n", + "75% 2.000000 1.000000 \n", + "max 5.000000 1.000000 " + ] + }, + "execution_count": 15, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Load demographic predictor variables\n", + "dem_predictors = [\"income\", \"age\", \"married\", \"kids\", \n", + " \"educ\", \"race\", \"lf\"] # lf = labor force status\n", + "\n", + "train_df[dem_predictors].describe()" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "id": "f4a1981e-d974-4788-b184-f4ab0d394a8c", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
housesequitytrustsstocksassetsavingwageinckginc
count2.310800e+042.310800e+042.310800e+042.310800e+042.310800e+042.310800e+042.310800e+042.310800e+04
mean8.474683e+053.559676e+066.433040e+051.371608e+061.606949e+076.660673e+041.664648e+054.352012e+05
std3.103222e+062.862459e+071.115652e+071.161028e+079.338054e+076.144518e+051.026294e+061.250478e+07
min0.000000e+000.000000e+000.000000e+000.000000e+000.000000e+000.000000e+000.000000e+00-1.030356e+06
25%0.000000e+000.000000e+000.000000e+000.000000e+005.828079e+040.000000e+000.000000e+000.000000e+00
50%2.086570e+051.124429e+040.000000e+000.000000e+004.228201e+052.318411e+014.720990e+040.000000e+00
75%5.680106e+053.263743e+050.000000e+000.000000e+002.134271e+061.159205e+041.262865e+050.000000e+00
max1.234322e+081.457828e+094.636821e+084.057219e+082.280388e+093.245775e+075.311114e+078.261732e+08
\n", + "
" + ], + "text/plain": [ + " houses equity trusts stocks asset \\\n", + "count 2.310800e+04 2.310800e+04 2.310800e+04 2.310800e+04 2.310800e+04 \n", + "mean 8.474683e+05 3.559676e+06 6.433040e+05 1.371608e+06 1.606949e+07 \n", + "std 3.103222e+06 2.862459e+07 1.115652e+07 1.161028e+07 9.338054e+07 \n", + "min 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 \n", + "25% 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 5.828079e+04 \n", + "50% 2.086570e+05 1.124429e+04 0.000000e+00 0.000000e+00 4.228201e+05 \n", + "75% 5.680106e+05 3.263743e+05 0.000000e+00 0.000000e+00 2.134271e+06 \n", + "max 1.234322e+08 1.457828e+09 4.636821e+08 4.057219e+08 2.280388e+09 \n", + "\n", + " saving wageinc kginc \n", + "count 2.310800e+04 2.310800e+04 2.310800e+04 \n", + "mean 6.660673e+04 1.664648e+05 4.352012e+05 \n", + "std 6.144518e+05 1.026294e+06 1.250478e+07 \n", + "min 0.000000e+00 0.000000e+00 -1.030356e+06 \n", + "25% 0.000000e+00 0.000000e+00 0.000000e+00 \n", + "50% 2.318411e+01 4.720990e+04 0.000000e+00 \n", + "75% 1.159205e+04 1.262865e+05 0.000000e+00 \n", + "max 3.245775e+07 5.311114e+07 8.261732e+08 " + ] + }, + "execution_count": 16, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Load financial predictor variables\n", + "fin_predictors = [\"houses\", \"equity\", \"trusts\", \"stocks\", \"asset\", \n", + " \"saving\", \"wageinc\", \"kginc\"]\n", + "\n", + "train_df[fin_predictors].describe()" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "id": "dd61b3c1-f0f8-4c7c-83ab-2885c771054b", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "count 2.310800e+04\n", + "mean 1.570575e+07\n", + "std 9.224071e+07\n", + "min -1.107621e+06\n", + "25% 2.445054e+04\n", + "50% 2.713352e+05\n", + "75% 1.872117e+06\n", + "max 2.280388e+09\n", + "Name: networth, dtype: float64" + ] + }, + "execution_count": 17, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "train_df[\"networth\"].describe()" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "id": "78b19005", + "metadata": {}, + "outputs": [], + "source": [ + "# predictors shared with cps data\n", + "\n", + "PREDICTORS = [\"hhsex\", # sex of head of household\n", + " \"age\", # age of respondent\n", + " \"married\", # marital status of respondent\n", + " \"kids\", # number of children in household\n", + " \"educ\", # highest level of education\n", + " \"race\", # race of respondent \n", + " \"income\", # total annual income of household \n", + " \"wageinc\", # income from wages and salaries\n", + " \"bussefarminc\", # income from business, self-employment or farm\n", + " \"intdivinc\", # income from interest and dividends\n", + " \"ssretinc\", # income from social security and retirement accounts\n", + " \"lf\", # labor force status\n", + " ] \n", + "\n", + "IMPUTED_VARIABLES = [\"networth\"] # some property also captured in cps data (HPROP_VAL)\n", + "\n", + "# additional predictors that may be useful for imputing wealth from scf data\n", + "\n", + "scf_imputed_variables = [\"networth\"]\n", + "\n", + "add_predictors = [\"houses\", \n", + " \"vehic\",\n", + " \"equity\", \n", + " \"trusts\", \n", + " \"stocks\", \n", + " \"asset\",\n", + " \"saving\",\n", + " ]" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "id": "307a6425", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
hhsexagemarriedkidseducraceincomewageincbussefarmincintdivincssretinclf
count23108.00000023108.00000023108.00000023108.00000023108.00000023108.0000002.310800e+042.310800e+042.310800e+042.310800e+042.310800e+0423108.000000
mean1.22407853.2386621.3767090.75125510.2520341.5486411.171661e+061.664648e+054.677982e+051.564668e+052.102781e+040.751471
std0.41698316.2253800.4845721.1274482.7080351.0568371.292835e+071.026294e+063.486120e+061.921528e+068.223668e+040.432169
min1.00000018.0000001.0000000.000000-1.0000001.0000000.000000e+000.000000e+00-2.360495e+060.000000e+000.000000e+000.000000
25%1.00000040.7500001.0000000.0000008.0000001.0000004.248891e+040.000000e+000.000000e+000.000000e+000.000000e+001.000000
50%1.00000054.0000001.0000000.00000011.0000001.0000009.264943e+044.720990e+040.000000e+000.000000e+000.000000e+001.000000
75%1.00000065.0000002.0000001.00000012.0000002.0000002.325088e+051.262865e+051.770371e+041.180247e+031.888396e+041.000000
max2.00000095.0000002.0000007.00000014.0000005.0000008.156336e+085.311114e+071.239260e+081.272071e+084.730904e+061.000000
\n", + "
" + ], + "text/plain": [ + " hhsex age married kids educ \\\n", + "count 23108.000000 23108.000000 23108.000000 23108.000000 23108.000000 \n", + "mean 1.224078 53.238662 1.376709 0.751255 10.252034 \n", + "std 0.416983 16.225380 0.484572 1.127448 2.708035 \n", + "min 1.000000 18.000000 1.000000 0.000000 -1.000000 \n", + "25% 1.000000 40.750000 1.000000 0.000000 8.000000 \n", + "50% 1.000000 54.000000 1.000000 0.000000 11.000000 \n", + "75% 1.000000 65.000000 2.000000 1.000000 12.000000 \n", + "max 2.000000 95.000000 2.000000 7.000000 14.000000 \n", + "\n", + " race income wageinc bussefarminc intdivinc \\\n", + "count 23108.000000 2.310800e+04 2.310800e+04 2.310800e+04 2.310800e+04 \n", + "mean 1.548641 1.171661e+06 1.664648e+05 4.677982e+05 1.564668e+05 \n", + "std 1.056837 1.292835e+07 1.026294e+06 3.486120e+06 1.921528e+06 \n", + "min 1.000000 0.000000e+00 0.000000e+00 -2.360495e+06 0.000000e+00 \n", + "25% 1.000000 4.248891e+04 0.000000e+00 0.000000e+00 0.000000e+00 \n", + "50% 1.000000 9.264943e+04 4.720990e+04 0.000000e+00 0.000000e+00 \n", + "75% 2.000000 2.325088e+05 1.262865e+05 1.770371e+04 1.180247e+03 \n", + "max 5.000000 8.156336e+08 5.311114e+07 1.239260e+08 1.272071e+08 \n", + "\n", + " ssretinc lf \n", + "count 2.310800e+04 23108.000000 \n", + "mean 2.102781e+04 0.751471 \n", + "std 8.223668e+04 0.432169 \n", + "min 0.000000e+00 0.000000 \n", + "25% 0.000000e+00 1.000000 \n", + "50% 0.000000e+00 1.000000 \n", + "75% 1.888396e+04 1.000000 \n", + "max 4.730904e+06 1.000000 " + ] + }, + "execution_count": 19, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "train_df[PREDICTORS].describe()" + ] + }, + { + "cell_type": "markdown", + "id": "8c130a96-c28d-4248-8571-e01df6186070", + "metadata": {}, + "source": [ + "## Imputing Wealth" + ] + }, + { + "cell_type": "code", + "execution_count": 20, + "id": "abbb6695-94ee-4972-9e0b-5444e83fc040", + "metadata": {}, + "outputs": [], + "source": [ + "import sys\n", + "import os\n", + "\n", + "# Get the parent directory of the current script\n", + "from policyengine_us_data.utils import QRF" + ] + }, + { + "cell_type": "code", + "execution_count": 21, + "id": "87906027-cd55-4375-8157-a4718dc500f1", + "metadata": {}, + "outputs": [], + "source": [ + "# Choose the imputed variable for our analysis\n", + "imputed_variable = [\"networth\"]\n", + "\n", + "y_train = train_df[imputed_variable]\n", + "y_test = test_df[imputed_variable]" + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "id": "01544abd-7d08-4050-832b-f3eed90b7bfe", + "metadata": {}, + "outputs": [], + "source": [ + "# Quantile loss function from https://colab.research.google.com/drive/1E8F7S1Uvfw_3PmpS226Sl1LWV5NBi0CE#scrollTo=cMzFV5QU5RZr \n", + "\n", + "def quantile_loss(q, y, f):\n", + " # q: Quantile to be evaluated, e.g., 0.5 for median.\n", + " # y: True value.\n", + " # f: Fitted or predicted value.\n", + " e = y - f\n", + " return np.maximum(q * e, (q - 1) * e)" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "id": "779dc98a-ed45-4a71-ba26-9f10b27d6977", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n" + ] + } + ], + "source": [ + "# Run a Quantile Regression Forest on demographic predictors\n", + "dem_qrf = QRF()\n", + "dem_qrf.fit(train_df[dem_predictors], train_df[imputed_variable])" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "id": "8d3a8bd6-b437-4393-9b9f-7a49a8d2f803", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n" + ] + } + ], + "source": [ + "# Run a Quantile Regression Forest on financial predictors\n", + "fin_qrf = QRF()\n", + "fin_qrf.fit(train_df[fin_predictors], train_df[imputed_variable])" + ] + }, + { + "cell_type": "markdown", + "id": "c83fe196-26db-471f-824f-751ce3c79d9e", + "metadata": {}, + "source": [ + "## Compare Results" + ] + }, + { + "cell_type": "code", + "execution_count": 25, + "id": "c27886f8-57ad-498b-b03a-3b62a9955f58", + "metadata": {}, + "outputs": [], + "source": [ + "QUANTILES = [0.05, 0.1, 0.3, 0.5, 0.7, 0.9, 0.95]\n", + "quantiles_legend = [str(int(q * 100)) + 'th percentile' for q in QUANTILES]" + ] + }, + { + "cell_type": "code", + "execution_count": 26, + "id": "d698661f-6bd2-40d5-8092-b0d3c7f6968b", + "metadata": {}, + "outputs": [], + "source": [ + "# Compute quantile loss for demographic predictors in train and test sets\n", + "dem_train_losses = [np.mean(quantile_loss(q, y_train.values.flatten(),\n", + " dem_qrf.predict(train_df[dem_predictors], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "dem_test_losses = [np.mean(quantile_loss(q, y_test.values.flatten(),\n", + " dem_qrf.predict(test_df[dem_predictors], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "\n", + "# Compute quantile loss for financial predictors in train and test sets\n", + "fin_train_losses = [np.mean(quantile_loss(q, y_train.values.flatten(),\n", + " fin_qrf.predict(train_df[fin_predictors], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "fin_test_losses = [np.mean(quantile_loss(q, y_test.values.flatten(),\n", + " fin_qrf.predict(test_df[fin_predictors], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "\n", + "# Create dataframes for train and test losses\n", + "dem_loss_df = pd.DataFrame([dem_train_losses, dem_test_losses], columns=quantiles_legend, index=[\"Train\", \"Test\"])\n", + "fin_loss_df = pd.DataFrame([fin_train_losses, fin_test_losses], columns=quantiles_legend, index=[\"Train\", \"Test\"])" + ] + }, + { + "cell_type": "code", + "execution_count": 27, + "id": "6ae17adb-ef57-4828-9cdb-a72b1f8743f1", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train483889.9829588.689220e+051.544056e+061.452119e+061.052197e+061.008559e+066.648588e+05
Test656693.3140461.125505e+062.899134e+063.153207e+062.810929e+061.774154e+061.097431e+06
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 483889.982958 8.689220e+05 1.544056e+06 1.452119e+06 \n", + "Test 656693.314046 1.125505e+06 2.899134e+06 3.153207e+06 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 1.052197e+06 1.008559e+06 6.648588e+05 \n", + "Test 2.810929e+06 1.774154e+06 1.097431e+06 " + ] + }, + "execution_count": 27, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "dem_loss_df" + ] + }, + { + "cell_type": "code", + "execution_count": 28, + "id": "1bf08ab2-3075-4e6b-ae40-bd82a562fe87", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train52081.506278108436.659586149232.359485146398.32186183133.39840358496.45722540373.180802
Test75691.390772139648.349920278392.896485244730.840757165099.31476977095.40261144476.467121
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 52081.506278 108436.659586 149232.359485 146398.321861 \n", + "Test 75691.390772 139648.349920 278392.896485 244730.840757 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 83133.398403 58496.457225 40373.180802 \n", + "Test 165099.314769 77095.402611 44476.467121 " + ] + }, + "execution_count": 28, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "fin_loss_df" + ] + }, + { + "cell_type": "code", + "execution_count": 29, + "id": "e5c8e37c-3cf7-4435-9cfd-15bb8f6e4da5", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plt.figure(figsize=(10,6))\n", + "plt.bar(dem_loss_df.columns, dem_loss_df.loc[\"Test\"], alpha=0.6, label=\"Demographic\", width=0.4, align='center', color='blue')\n", + "plt.bar(fin_loss_df.columns, fin_loss_df.loc[\"Test\"], alpha=0.6, label=\"Financial\", width=0.4, align='edge', color='green')\n", + "plt.title(\"Quantile Loss Comparison: Demographic vs Financial Predictors\", loc=\"left\")\n", + "sns.despine(left=True, bottom=True)\n", + "plt.xlabel(\"Quantile\")\n", + "plt.ylabel(\"Quantile Loss\")\n", + "plt.legend(title=\"Predictors\")\n", + "plt.xticks(rotation=45)\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 30, + "id": "8dc81333-f4b2-4560-b219-0eb5d4842be2", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "def plot_test_residuals(qrf, predictors, method, color):\n", + " # Predict different quantiles\n", + " q_10 = qrf.predict(test_df[predictors], mean_quantile=0.1)\n", + " q_50 = qrf.predict(test_df[predictors], mean_quantile=0.5)\n", + " q_90 = qrf.predict(test_df[predictors], mean_quantile=0.9)\n", + " \n", + " # Sort test samples for a smooth plot\n", + " sorted_idx = np.argsort(y_test.values.flatten())\n", + " y_sorted = y_test.iloc[sorted_idx].values.flatten()\n", + " \n", + " # Centered residuals (observed - predicted median)\n", + " residuals = y_sorted - q_50.iloc[sorted_idx].values.flatten()\n", + " \n", + " # Compute lower and upper prediction intervals (also centered)\n", + " lower_bound = q_10.iloc[sorted_idx].values.flatten() - q_50.iloc[sorted_idx].values.flatten()\n", + " upper_bound = q_90.iloc[sorted_idx].values.flatten() - q_50.iloc[sorted_idx].values.flatten()\n", + " \n", + " # Create the fan plot\n", + " plt.figure(figsize=(10,6))\n", + " plt.scatter(range(len(y_sorted)), residuals, color=color, alpha=0.6, s=10, label=\"Observed Residuals\")\n", + " plt.fill_between(range(len(y_sorted)), lower_bound, upper_bound, color='gray', alpha=0.5, label=\"Prediction Interval (Q=0.1 to Q=0.9)\")\n", + " plt.plot(range(len(y_sorted)), np.zeros_like(y_sorted), 'k--', label=\"Zero Error Line\")\n", + " plt.xlabel(\"Ordered Samples\")\n", + " plt.ylabel(\"Observed Values & Prediction Intervals (Centered)\")\n", + " plt.title(f\"Fan Plot of Predictions with Confidence Intervals for {method} Predictors\")\n", + " plt.legend()\n", + " plt.show()\n", + "\n", + "plot_test_residuals(dem_qrf, dem_predictors, \"Demographic\", \"blue\")" + ] + }, + { + "cell_type": "code", + "execution_count": 31, + "id": "54085664-3fd1-4b76-b297-2d5a083c2861", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_test_residuals(fin_qrf, fin_predictors, \"Financial\", \"green\")" + ] + }, + { + "cell_type": "code", + "execution_count": 32, + "id": "1ee7959c-f9b8-40b9-9749-a67f557c7831", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "def plot_test_predictions(qrf, predictors, method, color):\n", + " # Predict different quantiles\n", + " q_10 = qrf.predict(test_df[predictors], mean_quantile=0.1)\n", + " q_50 = qrf.predict(test_df[predictors], mean_quantile=0.5)\n", + " q_90 = qrf.predict(test_df[predictors], mean_quantile=0.9)\n", + " \n", + " # Sort test samples for a smooth plot\n", + " sorted_idx = np.argsort(y_test.values.flatten())\n", + " y_sorted = y_test.iloc[sorted_idx].values.flatten()\n", + " q_10_sorted = q_10.iloc[sorted_idx].values.flatten()\n", + " q_50_sorted = q_50.iloc[sorted_idx].values.flatten()\n", + " q_90_sorted = q_90.iloc[sorted_idx].values.flatten()\n", + " \n", + " # Plot the predicted Net Worth and its Confidence Intervals\n", + " plt.figure(figsize=(10,6))\n", + " plt.scatter(range(len(y_sorted)), y_sorted, color='red', alpha=0.75, s=10, label=\"Observed Net Worth\")\n", + " plt.fill_between(range(len(y_sorted)), q_10_sorted, q_90_sorted, color='gray', alpha=0.5, label=\"Prediction Interval (Q=0.1 to Q=0.9)\")\n", + " plt.scatter(range(len(y_sorted)), q_50_sorted, color=color, alpha=0.4, s=10, label=\"Predicted Net Worth (Q=0.5)\")\n", + " plt.xlabel(\"Ordered Samples\")\n", + " plt.ylabel(\"Observed & Predicted Net Worth\")\n", + " plt.title(f\"Net Worth Predictions with Confidence Intervals for {method} Predictors\")\n", + " plt.legend()\n", + " plt.show()\n", + "\n", + "plot_test_predictions(dem_qrf, dem_predictors, \"Demographic\", \"blue\")" + ] + }, + { + "cell_type": "code", + "execution_count": 33, + "id": "e602b3e4-182a-4139-900a-b1b90883dc45", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_test_predictions(fin_qrf, fin_predictors, \"Financial\", \"green\")" + ] + }, + { + "cell_type": "markdown", + "id": "e26c0791-5848-436a-a3ea-196018264e68", + "metadata": {}, + "source": [ + "## Benchmarking Methods for Survey of Consumer Finances Imputation\n", + "\n", + "Compared Methods: \n", + "\n", + "1. Quantile Regression Forests\n", + "2. Matching\n", + "3. OLS\n", + "4. QuantReg\n", + "5. Random Forests\n", + "6. GradientBoosting" + ] + }, + { + "cell_type": "markdown", + "id": "ceb50104", + "metadata": {}, + "source": [ + "### Preprocessing" + ] + }, + { + "cell_type": "code", + "execution_count": 51, + "id": "f2165b2b", + "metadata": {}, + "outputs": [], + "source": [ + "mean = train_df.mean(axis=0)\n", + "std = train_df.std(axis=0)\n", + "train_df = (train_df - mean) / std\n", + "test_df = (test_df - mean) / std" + ] + }, + { + "cell_type": "markdown", + "id": "5f59f2b4", + "metadata": {}, + "source": [ + "### Quantile Regression Forests" + ] + }, + { + "cell_type": "code", + "execution_count": 52, + "id": "d4609a28", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train459415.3925218.161909e+051.444721e+061.482116e+061.229121e+069.997076e+056.961144e+05
Test875513.2114271.211838e+062.266221e+062.493329e+062.184923e+061.682978e+061.096967e+06
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 459415.392521 8.161909e+05 1.444721e+06 1.482116e+06 \n", + "Test 875513.211427 1.211838e+06 2.266221e+06 2.493329e+06 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 1.229121e+06 9.997076e+05 6.961144e+05 \n", + "Test 2.184923e+06 1.682978e+06 1.096967e+06 " + ] + }, + "execution_count": 52, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# split data into train and test sets\n", + "train_df, test_df = train_test_split(data, test_size=0.2, train_size=0.8, random_state=42)\n", + "\n", + "y_train = train_df[IMPUTED_VARIABLES]\n", + "y_test = test_df[IMPUTED_VARIABLES]\n", + "\n", + "qrf = QRF()\n", + "qrf.fit(train_df[PREDICTORS], train_df[IMPUTED_VARIABLES])\n", + "\n", + "QUANTILES = [0.05, 0.1, 0.3, 0.5, 0.7, 0.9, 0.95]\n", + "quantiles_legend = [str(int(q * 100)) + 'th percentile' for q in QUANTILES]\n", + "\n", + "# compute quantile loss for demographic predictors in train and test sets\n", + "train_losses = [np.mean(quantile_loss(q, y_train.values.flatten(),\n", + " qrf.predict(train_df[PREDICTORS], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "test_losses = [np.mean(quantile_loss(q, y_test.values.flatten(),\n", + " qrf.predict(test_df[PREDICTORS], mean_quantile=q).values.flatten()\n", + " )) for q in QUANTILES]\n", + "\n", + "# create dataframes for train and test losses\n", + "loss_df = pd.DataFrame([train_losses, test_losses], columns=quantiles_legend, index=[\"Train\", \"Test\"])\n", + "\n", + "loss_df" + ] + }, + { + "cell_type": "code", + "execution_count": 59, + "id": "2bfd9837", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/sklearn/base.py:1389: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", + " return fit_method(estimator, *args, **kwargs)\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train508900.7729318.681492e+051.596949e+061.368721e+061.104342e+069.455935e+056.439730e+05
Test813647.2917311.197136e+062.431011e+062.552806e+062.232824e+061.684507e+061.115024e+06
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 508900.772931 8.681492e+05 1.596949e+06 1.368721e+06 \n", + "Test 813647.291731 1.197136e+06 2.431011e+06 2.552806e+06 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 1.104342e+06 9.455935e+05 6.439730e+05 \n", + "Test 2.232824e+06 1.684507e+06 1.115024e+06 " + ] + }, + "execution_count": 59, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# define number of folds\n", + "K = 5 \n", + "\n", + "quantile_losses = {\"train\": {q: [] for q in QUANTILES}, \"test\" : {q: [] for q in QUANTILES}}\n", + "\n", + "kf = KFold(n_splits=K, shuffle=True, random_state=42)\n", + "\n", + "for train_idx, test_idx in kf.split(data):\n", + " train_df, test_df = data.iloc[train_idx], data.iloc[test_idx]\n", + "\n", + " y_train = train_df[IMPUTED_VARIABLES]\n", + " y_test = test_df[IMPUTED_VARIABLES]\n", + "\n", + " qrf = QRF()\n", + " qrf.fit(train_df[PREDICTORS], train_df[IMPUTED_VARIABLES])\n", + "\n", + " # compute quantile loss for each quantile\n", + " for q in QUANTILES:\n", + " y_pred_train = qrf.predict(train_df[PREDICTORS], mean_quantile=q).values.flatten()\n", + " y_pred_test = qrf.predict(test_df[PREDICTORS], mean_quantile=q).values.flatten()\n", + " train_loss = np.mean(quantile_loss(q, y_train.values.flatten(), y_pred_train))\n", + " test_loss = np.mean(quantile_loss(q, y_test.values.flatten(), y_pred_test))\n", + " quantile_losses[\"train\"][q].append(train_loss)\n", + " quantile_losses[\"test\"][q].append(test_loss)\n", + "\n", + "# compute average quantile losses for train and test\n", + "avg_quantile_losses_train = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses[\"train\"][q]) for q in QUANTILES}\n", + "avg_quantile_losses_test = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses[\"test\"][q]) for q in QUANTILES}\n", + "\n", + "# create a dataframe to store both train and test losses\n", + "qrf_loss_df = pd.DataFrame([avg_quantile_losses_train, avg_quantile_losses_test], index=[\"Train\", \"Test\"])\n", + "\n", + "qrf_loss_df\n" + ] + }, + { + "cell_type": "code", + "execution_count": 54, + "id": "40b13081", + "metadata": {}, + "outputs": [], + "source": [ + "def plot_method_loss(method, loss_df, QUANTILES, K):\n", + " # extract quantile labels\n", + " quantile_labels = [f\"{int(q*100)}th\" for q in QUANTILES]\n", + "\n", + " # extract train and test losses\n", + " train_losses = loss_df.loc[\"Train\"].values\n", + " test_losses = loss_df.loc[\"Test\"].values\n", + "\n", + " # Set bar width and positions\n", + " x = np.arange(len(quantile_labels))\n", + " bar_width = 0.35\n", + "\n", + " # Create bar chart for train and test losses\n", + " plt.figure(figsize=(10, 6))\n", + " plt.bar(x - bar_width/2, train_losses, bar_width, label=\"Train Loss\", color='skyblue', edgecolor='black')\n", + " plt.bar(x + bar_width/2, test_losses, bar_width, label=\"Test Loss\", color='salmon', edgecolor='black')\n", + "\n", + " # labels and title\n", + " plt.xlabel(\"Quantile\", fontsize=12)\n", + " plt.ylabel(\"Average Quantile Loss\", fontsize=12)\n", + " plt.title(f\"Quantile Loss Across {K} Folds for {method}\", fontsize=14)\n", + " plt.xticks(x, quantile_labels)\n", + " plt.legend()\n", + " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", + "\n", + " plt.show()\n" + ] + }, + { + "cell_type": "code", + "execution_count": 60, + "id": "45362ea9", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_method_loss(\"Quantile Regression Forest\", qrf_loss_df, QUANTILES, K)" + ] + }, + { + "cell_type": "markdown", + "id": "4a929fd6", + "metadata": {}, + "source": [ + "## Matching" + ] + }, + { + "cell_type": "code", + "execution_count": 56, + "id": "d2785bfa", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "The rpy2.ipython extension is already loaded. To reload it, use:\n", + " %reload_ext rpy2.ipython\n", + "\n", + "The downloaded binary packages are in\n", + "\t/var/folders/yg/xdp70k_n4qj9ph1_0lm435c00000gp/T//RtmpvBN8NN/downloaded_packages\n" + ] + }, + { + "data": { + "text/plain": [ + "probando la URL 'https://cloud.r-project.org/bin/macosx/big-sur-arm64/contrib/4.4/StatMatch_1.4.3.tgz'\n", + "Content type 'application/x-gzip' length 593031 bytes (579 KB)\n", + "==================================================\n", + "downloaded 579 KB\n", + "\n" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "#!pip install rpy2 --upgrade\n", + "import rpy2\n", + "\n", + "# Install R package\n", + "%load_ext rpy2.ipython\n", + "%R install.packages('StatMatch', repos='https://cloud.r-project.org')\n", + "\n", + "from rpy2.robjects.packages import importr\n", + "from rpy2.robjects import pandas2ri\n", + "\n", + "# Enable R-Python DataFrame conversion\n", + "pandas2ri.activate()\n", + "StatMatch = importr(\"StatMatch\")" + ] + }, + { + "cell_type": "code", + "execution_count": 61, + "id": "3f07c389", + "metadata": {}, + "outputs": [], + "source": [ + "# split data into the same train and test sets\n", + "train_df, test_df = train_test_split(data, test_size=0.2, random_state=42)\n", + "train_df = data.loc[train_df.index] # Reassign correct indices\n", + "train_df = data.loc[train_df.index] # Reassign correct indices\n", + "\n", + "y_test = test_df[IMPUTED_VARIABLES]\n", + "\n", + "# normalize predictor variables\n", + "mean = train_df[PREDICTORS].mean(axis=0)\n", + "std = train_df[PREDICTORS].std(axis=0)\n", + "train_df[PREDICTORS] = (train_df[PREDICTORS] - mean) / std\n", + "test_df[PREDICTORS] = (test_df[PREDICTORS] - mean) / std\n", + "\n", + "# create a copy of the test set (to store imputed values later)\n", + "train_df_dup = train_df.copy()\n", + "test_df_dup = test_df.copy()\n", + "\n", + "# remove imputed variables from test_df before matching\n", + "test_df = test_df.drop(columns=IMPUTED_VARIABLES, errors=\"ignore\")\n", + "\n", + "# Function for Nearest Neighbor Hotdeck Matching\n", + "def nnd_hotdeck_using_rpy2(receiver, donor, matching_variables, z_variables, donor_classes=None):\n", + " \"\"\"\n", + " Perform statistical matching (nearest neighbor hotdeck imputation).\n", + " \"\"\"\n", + " assert receiver is not None and donor is not None\n", + " assert matching_variables is not None\n", + "\n", + " try:\n", + " if donor_classes:\n", + " out_NND = StatMatch.NND_hotdeck(\n", + " data_rec=receiver,\n", + " data_don=donor,\n", + " match_vars=pd.Series(matching_variables),\n", + " don_class=pd.Series(donor_classes),\n", + " k_donor=1\n", + " )\n", + " else:\n", + " out_NND = StatMatch.NND_hotdeck(\n", + " data_rec=receiver,\n", + " data_don=donor,\n", + " match_vars=pd.Series(matching_variables),\n", + " k_donor=1\n", + " )\n", + " except Exception as e:\n", + " print(\"Error in hotdeck matching:\", e)\n", + "\n", + " donor_idx = np.array(out_NND[0], dtype=int)\n", + " receiver_idx = np.array(receiver.index, dtype=int)\n", + "\n", + " if len(donor_idx) != len(receiver_idx):\n", + " raise ValueError(\"Mismatch between donor_idx and test_df lengths. \"\n", + " f\"{len(donor_idx)} vs {len(receiver_idx)}\")\n", + "\n", + " # Create the Nx2 \"matching ID\" matrix that create_fused() expects\n", + " mtc_ids = np.column_stack((receiver_idx, donor_idx))\n", + "\n", + " # Now call create_fused with mtc_ids\n", + " fused_0 = StatMatch.create_fused(\n", + " data_rec=receiver, # test_df\n", + " data_don=donor, # train_df\n", + " mtc_ids=mtc_ids,\n", + " z_vars=pd.Series(z_variables)\n", + " )\n", + "\n", + " # Convert R dataframe to Pandas dataframe\n", + " return pandas2ri.rpy2py_dataframe(fused_0)\n" + ] + }, + { + "cell_type": "code", + "execution_count": 58, + "id": "a490dfe1", + "metadata": {}, + "outputs": [ + { + "ename": "ValueError", + "evalue": "Mismatch between donor_idx and test_df lengths. 11554 vs 5777", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mValueError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[58], line 2\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[38;5;66;03m# Perform statistical matching\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m fused0 \u001b[38;5;241m=\u001b[39m nnd_hotdeck_using_rpy2(\n\u001b[1;32m 3\u001b[0m receiver\u001b[38;5;241m=\u001b[39mtest_df,\n\u001b[1;32m 4\u001b[0m donor\u001b[38;5;241m=\u001b[39mtrain_df,\n\u001b[1;32m 5\u001b[0m matching_variables\u001b[38;5;241m=\u001b[39mPREDICTORS,\n\u001b[1;32m 6\u001b[0m z_variables\u001b[38;5;241m=\u001b[39mIMPUTED_VARIABLES,\n\u001b[1;32m 7\u001b[0m donor_classes\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m 8\u001b[0m )\n", + "Cell \u001b[0;32mIn[57], line 52\u001b[0m, in \u001b[0;36mnnd_hotdeck_using_rpy2\u001b[0;34m(receiver, donor, matching_variables, z_variables, donor_classes)\u001b[0m\n\u001b[1;32m 49\u001b[0m receiver_idx \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39marray(receiver\u001b[38;5;241m.\u001b[39mindex, dtype\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mint\u001b[39m)\n\u001b[1;32m 51\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(donor_idx) \u001b[38;5;241m!=\u001b[39m \u001b[38;5;28mlen\u001b[39m(receiver_idx):\n\u001b[0;32m---> 52\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mMismatch between donor_idx and test_df lengths. \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 53\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mlen\u001b[39m(donor_idx)\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m vs \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mlen\u001b[39m(receiver_idx)\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m 55\u001b[0m \u001b[38;5;66;03m# Create the Nx2 \"matching ID\" matrix that create_fused() expects\u001b[39;00m\n\u001b[1;32m 56\u001b[0m mtc_ids \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39mcolumn_stack((receiver_idx, donor_idx))\n", + "\u001b[0;31mValueError\u001b[0m: Mismatch between donor_idx and test_df lengths. 11554 vs 5777" + ] + } + ], + "source": [ + "# Perform statistical matching\n", + "fused0 = nnd_hotdeck_using_rpy2(\n", + " receiver=test_df,\n", + " donor=train_df,\n", + " matching_variables=PREDICTORS,\n", + " z_variables=IMPUTED_VARIABLES,\n", + " donor_classes=None\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6720fd81", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " method q x label pred\n", + "0 Matching 0.05 9199 -68856.794702 NaN\n", + "1 Matching 0.10 9199 -68856.794702 NaN\n", + "2 Matching 0.30 9199 -68856.794702 NaN\n", + "3 Matching 0.50 9199 -68856.794702 NaN\n", + "4 Matching 0.70 9199 -68856.794702 NaN\n" + ] + } + ], + "source": [ + "# Store imputed values in preds DataFrame for evaluation\n", + "preds = pd.DataFrame({\n", + " 'method': 'Matching',\n", + " 'q': np.tile(QUANTILES, len(test_df)),\n", + " 'x': np.repeat(test_df.index, len(QUANTILES)),\n", + " 'label': np.repeat(y_test.values.flatten(), len(QUANTILES)),\n", + " 'pred': np.repeat(fused0[IMPUTED_VARIABLES].values.flatten(), len(QUANTILES))\n", + "})\n", + "\n", + "# Display first few rows\n", + "print(preds.head())" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ce66382a", + "metadata": {}, + "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "622200f3", + "metadata": {}, + "outputs": [], + "source": [] + }, + { + "cell_type": "markdown", + "id": "b23adae9", + "metadata": {}, + "source": [ + "### OLS" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "56510bb8", + "metadata": {}, + "outputs": [], + "source": [ + "import statsmodels.api as sm\n", + "from scipy.stats import norm" + ] + }, + { + "cell_type": "code", + "execution_count": 62, + "id": "7ea12679", + "metadata": {}, + "outputs": [], + "source": [ + "def ols_quantile(m, X, q):\n", + " # m: OLS model.\n", + " # X: X matrix.\n", + " # q: Quantile.\n", + " #\n", + " # Set alpha based on q. Vectorized for different values of q.\n", + " mean_pred = m.predict(X)\n", + " se = np.sqrt(m.scale)\n", + " return mean_pred + norm.ppf(q) * se" + ] + }, + { + "cell_type": "code", + "execution_count": 63, + "id": "52f33de6", + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train8.035239e+076.319842e+072.800911e+071.129672e+073.275195e+076.721998e+078.401729e+07
Test8.046019e+076.331117e+072.813116e+071.141502e+073.286725e+076.733995e+078.413717e+07
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 8.035239e+07 6.319842e+07 2.800911e+07 1.129672e+07 \n", + "Test 8.046019e+07 6.331117e+07 2.813116e+07 1.141502e+07 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 3.275195e+07 6.721998e+07 8.401729e+07 \n", + "Test 3.286725e+07 6.733995e+07 8.413717e+07 " + ] + }, + "execution_count": 63, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "quantile_losses_ols = {\"train\": {q: [] for q in QUANTILES}, \"test\": {q: [] for q in QUANTILES}}\n", + "\n", + "for train_idx, test_idx in kf.split(data):\n", + " train_df, test_df = data.iloc[train_idx], data.iloc[test_idx]\n", + "\n", + " y_train = train_df[IMPUTED_VARIABLES]\n", + " y_test = test_df[IMPUTED_VARIABLES]\n", + "\n", + " # add constant for OLS\n", + " x_train = sm.add_constant(train_df[PREDICTORS])\n", + " x_test = sm.add_constant(test_df[PREDICTORS])\n", + "\n", + " # fit OLS model\n", + " ols_model = sm.OLS(y_train, x_train).fit()\n", + "\n", + " # compute quantile loss for each quantile\n", + " for q in QUANTILES:\n", + " y_pred_train = ols_quantile(ols_model, x_train, q)\n", + " y_pred_test = ols_quantile(ols_model, x_test, q)\n", + " train_loss = np.mean(np.abs(y_train.values.flatten() - y_pred_train.values.flatten()))\n", + " test_loss = np.mean(np.abs(y_test.values.flatten() - y_pred_test.values.flatten()))\n", + " quantile_losses_ols[\"train\"][q].append(train_loss)\n", + " quantile_losses_ols[\"test\"][q].append(test_loss)\n", + "\n", + "# compute average quantile losses for train and test\n", + "avg_quantile_losses_train_ols = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses_ols[\"train\"][q]) for q in QUANTILES}\n", + "avg_quantile_losses_test_ols = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses_ols[\"test\"][q]) for q in QUANTILES}\n", + "\n", + "# create a dataframe to store both train and test losses\n", + "ols_loss_df = pd.DataFrame([avg_quantile_losses_train_ols, avg_quantile_losses_test_ols], index=[\"Train\", \"Test\"])\n", + "\n", + "ols_loss_df" + ] + }, + { + "cell_type": "code", + "execution_count": 64, + "id": "a0efa258", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_method_loss(\"OLS\", ols_loss_df, QUANTILES, K)" + ] + }, + { + "cell_type": "markdown", + "id": "9b3dfd0e", + "metadata": {}, + "source": [ + "### QuantReg" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "46c64f23", + "metadata": {}, + "outputs": [], + "source": [ + "quantreg_full = sm.QuantReg(y_train, x_train)\n", + "\n", + "preds.loc[preds.method == 'QuantReg', 'pred'] = np.concatenate(\n", + " [quantreg_full.fit(q=q).predict(x_test) for q in QUANTILES])" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e5dd03da", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n", + "/Users/movil1/anaconda3/lib/python3.11/site-packages/statsmodels/regression/quantile_regression.py:191: IterationLimitWarning: Maximum number of iterations (1000) reached.\n", + " warnings.warn(\"Maximum number of iterations (\" + str(max_iter) +\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
5th percentile10th percentile30th percentile50th percentile70th percentile90th percentile95th percentile
Train1.194896e+071.090204e+078.867122e+068.109575e+069.039018e+061.746481e+072.425136e+07
Test1.193932e+071.094884e+078.963877e+068.344511e+069.153171e+061.803404e+072.438265e+07
\n", + "
" + ], + "text/plain": [ + " 5th percentile 10th percentile 30th percentile 50th percentile \\\n", + "Train 1.194896e+07 1.090204e+07 8.867122e+06 8.109575e+06 \n", + "Test 1.193932e+07 1.094884e+07 8.963877e+06 8.344511e+06 \n", + "\n", + " 70th percentile 90th percentile 95th percentile \n", + "Train 9.039018e+06 1.746481e+07 2.425136e+07 \n", + "Test 9.153171e+06 1.803404e+07 2.438265e+07 " + ] + }, + "execution_count": 66, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "quantile_losses_qr = {\"train\": {q: [] for q in QUANTILES}, \"test\": {q: [] for q in QUANTILES}}\n", + "\n", + "for train_idx, test_idx in kf.split(data):\n", + " train_df, test_df = data.iloc[train_idx], data.iloc[test_idx]\n", + "\n", + " y_train = train_df[IMPUTED_VARIABLES]\n", + " y_test = test_df[IMPUTED_VARIABLES]\n", + "\n", + " # add constant for QuantReg\n", + " x_train = sm.add_constant(train_df[PREDICTORS])\n", + " x_test = sm.add_constant(test_df[PREDICTORS])\n", + "\n", + " # fit and predict for each quantile\n", + " for q in QUANTILES:\n", + " quantreg_model = sm.QuantReg(y_train, x_train).fit(q=q)\n", + " y_pred_train = quantreg_model.predict(x_train)\n", + " y_pred_test = quantreg_model.predict(x_test)\n", + "\n", + " # compute quantile loss\n", + " train_loss = np.mean(np.abs(y_train.values.flatten() - y_pred_train.values.flatten()))\n", + " test_loss = np.mean(np.abs(y_test.values.flatten() - y_pred_test.values.flatten()))\n", + " quantile_losses_qr[\"train\"][q].append(train_loss)\n", + " quantile_losses_qr[\"test\"][q].append(test_loss)\n", + "\n", + "# compute average quantile losses for train and test\n", + "avg_quantile_losses_train_qr = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses_qr[\"train\"][q]) for q in QUANTILES}\n", + "avg_quantile_losses_test_qr = {f\"{int(q*100)}th percentile\": np.mean(quantile_losses_qr[\"test\"][q]) for q in QUANTILES}\n", + "\n", + "# create a dataframe to store both train and test losses for QuantReg\n", + "quantreg_loss_df = pd.DataFrame([avg_quantile_losses_train_qr, avg_quantile_losses_test_qr], index=[\"Train\", \"Test\"])\n", + "\n", + "quantreg_loss_df" + ] + }, + { + "cell_type": "code", + "execution_count": 67, + "id": "40b87f78", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_method_loss(\"QuantReg\", quantreg_loss_df, QUANTILES, K)" + ] + }, + { + "cell_type": "markdown", + "id": "89276e9e", + "metadata": {}, + "source": [ + "### Random Forests" + ] + }, + { + "cell_type": "markdown", + "id": "32a58fb1", + "metadata": {}, + "source": [ + "### Gradient Boosting" + ] + }, + { + "cell_type": "markdown", + "id": "84010b08", + "metadata": {}, + "source": [ + "### Comparing Across Methods" + ] + }, + { + "cell_type": "code", + "execution_count": 73, + "id": "a05ce574", + "metadata": {}, + "outputs": [], + "source": [ + "quantreg_loss_df[\"Method\"] = \"QuantReg\"\n", + "ols_loss_df[\"Method\"] = \"OLS\"\n", + "qrf_loss_df[\"Method\"] = \"QRF\"\n", + "\n", + "# Convert each DataFrame from wide to long format\n", + "quantreg_long = quantreg_loss_df.reset_index().melt(id_vars=[\"index\", \"Method\"], var_name=\"Quantile\", value_name=\"Loss\")\n", + "ols_long = ols_loss_df.reset_index().melt(id_vars=[\"index\", \"Method\"], var_name=\"Quantile\", value_name=\"Loss\")\n", + "qrf_long = qrf_loss_df.reset_index().melt(id_vars=[\"index\", \"Method\"], var_name=\"Quantile\", value_name=\"Loss\")\n", + "\n", + "# Rename the 'index' column to 'Split' (Train/Test)\n", + "for df in [quantreg_long, ols_long, qrf_long]:\n", + " df.rename(columns={\"index\": \"Split\"}, inplace=True)\n", + "\n", + "# Concatenate all into a single DataFrame\n", + "combined_loss_df = pd.concat([quantreg_long, ols_long, qrf_long], ignore_index=True)" + ] + }, + { + "cell_type": "code", + "execution_count": 72, + "id": "b91edc3c", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "test_loss_df = combined_loss_df[combined_loss_df[\"Split\"] == \"Test\"]\n", + "\n", + "plt.figure(figsize=(12, 6))\n", + "sns.barplot(data=test_loss_df, x=\"Quantile\", y=\"Loss\", hue=\"Method\", dodge=True)\n", + "plt.xlabel(\"Quantile\", fontsize=12)\n", + "plt.ylabel(\"Average Test Quantile Loss\", fontsize=12)\n", + "plt.title(\"Test Loss Across Quantiles for Different Imputation Methods\", fontsize=14)\n", + "plt.legend(title=\"Method\")\n", + "plt.xticks(rotation=45)\n", + "plt.grid(axis='y', linestyle='--', alpha=0.7)\n", + "plt.show()" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "base", + "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": "3.11.8" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/docs/scf/imputing_scf.md b/docs/scf/imputing_scf.md new file mode 100644 index 00000000..e69de29b