Skip to content

Instantly share code, notes, and snippets.

@fonnesbeck

fonnesbeck/Technology Paper.ipynb

Created September 16, 2015 15:55
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save fonnesbeck/45b9a09260969d853949 to your computer and use it in GitHub Desktop.
Save fonnesbeck/45b9a09260969d853949 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
Technology Paper.ipynb
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Technology Effects Analysis"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# Import modules and set options\n",
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"pd.set_option('display.notebook_repr_html', False)\n",
"pd.set_option('display.max_columns', 20)\n",
"pd.set_option('display.max_rows', 100)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Import data"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python3.4/site-packages/pandas/io/parsers.py:1170: DtypeWarning: Columns (31,33,41) have mixed types. Specify dtype option on import or set low_memory=False.\n",
" data = self._reader.read(nrows)\n"
]
}
],
"source": [
"lsl_dr = pd.read_csv('lsl_dr.csv', index_col=0)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"0 initial_assessment_arm_1 2009-2010 0 0 0 0 0 \n",
"1 initial_assessment_arm_1 2010-2011 0 0 7 0 2 \n",
"2 initial_assessment_arm_1 2010-2011 0 0 7 0 2 \n",
"3 initial_assessment_arm_1 2009-2010 0 1 7 0 1 \n",
"4 initial_assessment_arm_1 2005-2006 0 1 0 0 3 \n",
"\n",
" _mother_ed father_ed premature_age ... known_synd \\\n",
"0 5 4 9 ... 0 \n",
"1 6 6 7 ... 0 \n",
"2 6 6 7 ... 1 \n",
"3 6 6 8 ... 0 \n",
"4 2 2 8 ... 0 \n",
"\n",
" synd_or_disab race age_test domain school score test_name \\\n",
"0 1 0 NaN NaN NaN NaN NaN \n",
"1 0 NaN NaN NaN NaN NaN NaN \n",
"2 1 NaN NaN NaN NaN NaN NaN \n",
"3 0 NaN NaN NaN NaN NaN NaN \n",
"4 NaN 0 48 Articulation 521 41 NaN \n",
"\n",
" test_type academic_year_start \n",
"0 NaN 2009 \n",
"1 NaN 2010 \n",
"2 NaN 2010 \n",
"3 NaN 2009 \n",
"4 Goldman 2005 \n",
"\n",
"[5 rows x 74 columns]"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lsl_dr.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Indicator for non-profound hearing loss"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['deg_hl_below6'] = lsl_dr.degree_hl<6\n",
"lsl_dr.loc[lsl_dr.degree_hl.isnull(), 'deg_hl_below6'] = np.nan"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Indicator for first intervention outside OPTION"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['int_outside_option'] = lsl_dr.age > lsl_dr.age_int\n",
"lsl_dr.loc[lsl_dr.age < lsl_dr.age_int, 'int_outside_option'] = np.nan"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Indicator for high school graduation of mother"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['mother_hs'] = lsl_dr.mother_ed > 1\n",
"lsl_dr.loc[lsl_dr.mother_ed.isnull(), 'mother_hs'] = None"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Joint commission benchmark"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([ 1. , 18. , 17. , 22. , nan, 3. , 2. , 5. ,\n",
" 13. , 36. , 28. , 21. , 41. , 26. , 0. , 67. ,\n",
" 20. , 24. , 4. , 40. , 60. , 10. , 6. , 25. ,\n",
" 7. , 27. , 15. , 35. , 14. , 42. , 34. , 12. ,\n",
" 9. , 32. , 50. , 8. , 33. , 23. , 11. , 31. ,\n",
" 30. , 49. , 48. , 1.5, 19. , 2.5, 39. , 52. ,\n",
" 16. , 38. , 29. , 51. , 46. , 45. , 54. , 88. ,\n",
" 65. , 44. , 81. , 116. , 72. , 57. , 62. , 43. ,\n",
" 78. , 83. , 61. , 107. , 64. , 74. , 37. , 77. ,\n",
" 96. , 97. , 79. , 47. , 53. , 59. , 84. , 95. ,\n",
" 80. , 0.5, 58. , 56. , 86. , 98. , 85. , 75. ,\n",
" 119. , 66. , 70. , 63. , 140. , 126. , 133. , 103. ,\n",
" 87. , 76. , 55. , 68. , 92. , 71. , 154. , 89. ,\n",
" 152. ])"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lsl_dr.onset_1.unique()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['ident_by_3'] = ident_by_3 = lsl_dr.onset_1 <= 3\n",
"lsl_dr['program_by_6'] = program_by_6 = lsl_dr.age <= 6\n",
"lsl_dr['jcih'] = ident_by_3 & program_by_6\n",
"lsl_dr.loc[lsl_dr.onset_1.isnull() | lsl_dr.age.isnull(), 'jcih'] = None"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Create age in years variable"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['age_years'] = lsl_dr.age/12."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Retain only 4-year-olds"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr = lsl_dr[(lsl_dr.age_test>=48) & (lsl_dr.age_test<60)]"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(6873, 81)"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lsl_dr.shape"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(1613,)"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lsl_dr.study_id.unique().shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Summary Statistics for Paper"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"unique_students = lsl_dr.drop_duplicates(subset='study_id')"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(1613, 81)"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.shape"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1593.000000\n",
"mean 29.700565\n",
"std 15.676539\n",
"min 1.000000\n",
"25% 18.000000\n",
"50% 32.000000\n",
"75% 41.000000\n",
"max 68.000000\n",
"Name: age, dtype: float64"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.age.describe()"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.5236612702366127"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.male.mean()"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"1 841\n",
"0 765\n",
"dtype: int64"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.male.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1501.000000\n",
"mean 1.161226\n",
"std 0.928434\n",
"min 0.000000\n",
"25% 0.000000\n",
"50% 1.000000\n",
"75% 2.000000\n",
"max 3.000000\n",
"Name: sib, dtype: float64"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.sib.describe()"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1613\n",
"mean 0.4575325\n",
"std 0.4983478\n",
"min False\n",
"25% 0\n",
"50% 0\n",
"75% 1\n",
"max True\n",
"Name: degree_hl, dtype: object"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"(unique_students.degree_hl == 6).describe()"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"220"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.deg_hl_below6.isnull().sum()"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0 738\n",
"1 655\n",
"dtype: int64"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.deg_hl_below6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1613\n",
"mean 0.3787973\n",
"std 0.485238\n",
"min False\n",
"25% 0\n",
"50% 0\n",
"75% 1\n",
"max True\n",
"Name: degree_hl, dtype: object"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"((unique_students.degree_hl < 6) & (unique_students.degree_hl > 1)).describe()"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1006.000000\n",
"mean 0.580517\n",
"std 0.493720\n",
"min 0.000000\n",
"25% 0.000000\n",
"50% 1.000000\n",
"75% 1.000000\n",
"max 1.000000\n",
"Name: mother_hs, dtype: float64"
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.mother_hs.describe()"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"count 1413.000000\n",
"mean 0.213022\n",
"std 0.409588\n",
"min 0.000000\n",
"25% 0.000000\n",
"50% 0.000000\n",
"75% 0.000000\n",
"max 1.000000\n",
"Name: synd_or_disab, dtype: float64"
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.synd_or_disab.describe()"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"male \n",
"0 False 670\n",
" True 95\n",
"1 False 750\n",
" True 91\n",
"dtype: int64"
]
},
"execution_count": 25,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('male').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"sib \n",
"0 False 346\n",
" True 38\n",
"1 False 562\n",
" True 90\n",
"2 False 267\n",
" True 37\n",
"3 False 145\n",
" True 16\n",
"dtype: int64"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('sib').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"deg_hl_below6 \n",
"0 False 638\n",
" True 100\n",
"1 False 574\n",
" True 81\n",
"dtype: int64"
]
},
"execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('deg_hl_below6').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"mother_hs \n",
"0 False 367\n",
" True 55\n",
"1 False 489\n",
" True 95\n",
"dtype: int64"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('mother_hs').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"synd_or_disab \n",
"0 False 976\n",
" True 136\n",
"1 False 273\n",
" True 28\n",
"dtype: int64"
]
},
"execution_count": 29,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('synd_or_disab').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"ident_by_3 \n",
"False False 1013\n",
" True 34\n",
"True False 414\n",
" True 152\n",
"dtype: int64"
]
},
"execution_count": 30,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('ident_by_3').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"jcih \n",
"0 False 928\n",
" True 17\n",
"1 True 152\n",
"dtype: int64"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"unique_students.groupby('jcih').program_by_6.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEQCAYAAACwSgOGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIEiSCgaCJAmA/w+nsoeAk7rRaQAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10d2a34e0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"grouped_age = unique_students.replace({'jcih':{1:'JCIH', 0:'non-JCIH'}}).groupby('jcih')['age_years']\n",
"fig, ax = plt.subplots()\n",
"\n",
"for k, v in grouped_age:\n",
" v.hist(label=k, alpha=.75, ax=ax, grid=False)\n",
"\n",
"ax.legend()\n",
"plt.xlabel('Age (years)'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"students_data = lsl_dr.copy()"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"students_data['school_id'] = students_data.school.replace({int(s):i for i,s in enumerate(students_data.school.unique())})"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"students_data = students_data[students_data.domain=='Language']"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(2594, 82)"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"students_data.shape"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"option_scores = students_data.drop_duplicates('study_id', take_last=True)[['score', 'school_id', \n",
" 'deg_hl_below6', 'mother_hs']].dropna()"
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"option_scores = option_scores.astype(int)"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"option_scores.to_csv('option_scores.csv', index=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Create technology indicators"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"unimodal = ~lsl_dr[[u'bilateral_ci',\n",
" u'bilateral_ha', \n",
" u'bimodal']].sum(1).astype(bool)"
]
},
{
"cell_type": "code",
"execution_count": 41,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"lsl_dr['unimodal_ci'] = unimodal & lsl_dr.cochlear\n",
"lsl_dr['unimodal_ha'] = unimodal & lsl_dr.hearing_aid"
]
},
{
"cell_type": "code",
"execution_count": 42,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"bilateral_ci 2136\n",
"bilateral_ha 2261\n",
"bimodal 791\n",
"unimodal_ci 311\n",
"unimodal_ha 320\n",
"dtype: int64"
]
},
"execution_count": 42,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lsl_dr[[u'bilateral_ci',\n",
" u'bilateral_ha', \n",
" u'bimodal', 'unimodal_ci', 'unimodal_ha']].sum(0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Since bilateral hearing aid is the most prevalent category, it will be used as the baseline category."
]
},
{
"cell_type": "code",
"execution_count": 43,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"tech_index = lsl_dr[[u'bilateral_ci',\n",
" u'bilateral_ha', \n",
" u'bimodal', 'unimodal_ci', 'unimodal_ha']].sum(1).astype(bool)\n",
"\n",
"technology_subset = lsl_dr[tech_index]"
]
},
{
"cell_type": "code",
"execution_count": 44,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"146.0"
]
},
"execution_count": 44,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"technology_subset.score.max()"
]
},
{
"cell_type": "code",
"execution_count": 45,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"572"
]
},
"execution_count": 45,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"((unique_students.cochlear>0) & (unique_students.age_amp.notnull())).sum()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Receptive Language Test Score Model"
]
},
{
"cell_type": "code",
"execution_count": 46,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"6 initial_assessment_arm_1 2005-2006 0 1 0 0 3 \n",
"250 initial_assessment_arm_1 2012-2013 0 1 0 0 2 \n",
"272 initial_assessment_arm_1 2014-2015 0 0 2 0 2 \n",
"293 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"302 initial_assessment_arm_1 2011-2012 0 1 0 0 2 \n",
"\n",
" _mother_ed father_ed premature_age ... academic_year_start \\\n",
"6 2 2 8 ... 2005 \n",
"250 4 6 7 ... 2012 \n",
"272 6 6 8 ... 2014 \n",
"293 5 6 7 ... 2014 \n",
"302 6 6 8 ... 2011 \n",
"\n",
" deg_hl_below6 int_outside_option mother_hs ident_by_3 program_by_6 \\\n",
"6 0 1 0 False False \n",
"250 1 1 1 False False \n",
"272 1 1 NaN False False \n",
"293 1 1 1 False False \n",
"302 0 1 NaN False False \n",
"\n",
" jcih age_years unimodal_ci unimodal_ha \n",
"6 0 4.500000 False False \n",
"250 0 4.666667 True False \n",
"272 0 3.583333 False False \n",
"293 0 4.083333 False False \n",
"302 0 4.416667 False False \n",
"\n",
"[5 rows x 83 columns]"
]
},
"execution_count": 46,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Filter domain and English-only, and appropriate bilateral subset\n",
"receptive_language_dataset = technology_subset[(technology_subset.domain=='Language') \n",
" & (technology_subset.test_type=='receptive') \n",
" & (technology_subset.non_english==False)]\n",
"\n",
"receptive_language_dataset.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This is the size of the resulting dataset to be used in this analysis"
]
},
{
"cell_type": "code",
"execution_count": 47,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(941, 83)"
]
},
"execution_count": 47,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_language_dataset.shape"
]
},
{
"cell_type": "code",
"execution_count": 48,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mean_score = receptive_language_dataset.groupby('study_id')['score'].mean()"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset = receptive_language_dataset.drop_duplicates('study_id')"
]
},
{
"cell_type": "code",
"execution_count": 50,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset = receptive_language_dataset.set_index('study_id')"
]
},
{
"cell_type": "code",
"execution_count": 51,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset['mean_score'] = mean_score"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Specify criterion to use: `jcih`, `ident_by_3`, `program_by_6`"
]
},
{
"cell_type": "code",
"execution_count": 52,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"criterion_labels = {'jcih': 'JCIH',\n",
" 'ident_by_3': 'Diagnosed by 3 months',\n",
" 'program_by_6': 'Intervention by 6 months'}\n",
"\n",
"criterion = 'jcih'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Proportion of missing values for each variable"
]
},
{
"cell_type": "code",
"execution_count": 53,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"age_years 0.00\n",
"school 0.00\n",
"score 0.00\n",
"male 0.00\n",
"time 0.04\n",
"family_inv 0.16\n",
"sib 0.06\n",
"deg_hl_below6 0.01\n",
"mother_hs 0.33\n",
"synd_or_disab 0.14\n",
"jcih 0.20\n",
"dtype: float64"
]
},
"execution_count": 53,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"covs = ['age_years', 'school', 'score', 'male', 'time',\n",
" 'family_inv', 'sib', 'deg_hl_below6', 'mother_hs', 'synd_or_disab'] + [criterion]\n",
"(receptive_language_dataset[covs].isnull().sum() / receptive_language_dataset.shape[0]).round(2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Drop individuals with missing `age`, since there is <1% of them"
]
},
{
"cell_type": "code",
"execution_count": 54,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset = receptive_language_dataset[receptive_language_dataset.age.notnull()]"
]
},
{
"cell_type": "code",
"execution_count": 55,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAF/CAYAAADn6NV5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10cd98a20>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"receptive_language_dataset.score.hist(grid=False, bins=25, figsize=(10, 6))\n",
"plt.xlabel('Standard Scores'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 56,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x10c0d9ac8>"
]
},
"execution_count": 56,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEBCAYAAACZhwWsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngAZQ9gDQAMoeABrw/9C31YaZ6otzAAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10d295e80>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"receptive_language_dataset.mother_ed.hist()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Final analysis dataset size"
]
},
{
"cell_type": "code",
"execution_count": 57,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(871, 83)"
]
},
"execution_count": 57,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_language_dataset.shape"
]
},
{
"cell_type": "code",
"execution_count": 58,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"mother_hs 33.3\n",
"jcih 19.3\n",
"family_inv 15.8\n",
"synd_or_disab 14.0\n",
"sib 5.5\n",
"time 3.6\n",
"deg_hl_below6 1.3\n",
"male 0.0\n",
"score 0.0\n",
"school 0.0\n",
"age_years 0.0\n",
"dtype: float64"
]
},
"execution_count": 58,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"missing_pct = (receptive_language_dataset[covs].isnull().mean()*100).round(1)\n",
"missing_pct.sort(ascending=False)\n",
"missing_pct"
]
},
{
"cell_type": "code",
"execution_count": 59,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset['age_amp_missing'] = receptive_language_dataset.age_amp.isnull()\n",
"by_age_amp_missing = receptive_language_dataset.groupby('age_amp_missing')"
]
},
{
"cell_type": "code",
"execution_count": 60,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" male synd_or_disab deg_hl_below6 mother_hs\n",
"age_amp_missing \n",
"False 0.510885 0.221106 0.469298 0.616034\n",
"True 0.478022 0.250000 0.386364 0.654206"
]
},
"execution_count": 60,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"by_age_amp_missing['male', 'synd_or_disab', 'deg_hl_below6', 'mother_hs',\n",
" 'non_english'].mean()"
]
},
{
"cell_type": "code",
"execution_count": 61,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_language_dataset['jcih_missing'] = receptive_language_dataset.jcih.isnull()\n",
"by_jcih_missing = receptive_language_dataset.groupby('jcih_missing')"
]
},
{
"cell_type": "code",
"execution_count": 62,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" male synd_or_disab deg_hl_below6 mother_hs\n",
"jcih_missing \n",
"False 0.513514 0.219110 0.467811 0.626033\n",
"True 0.464286 0.260563 0.385093 0.608247"
]
},
"execution_count": 62,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"by_jcih_missing['male', 'synd_or_disab', 'deg_hl_below6', 'mother_hs',\n",
" 'non_english'].mean()"
]
},
{
"cell_type": "code",
"execution_count": 63,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"2.2484691924990403"
]
},
"execution_count": 63,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_language_dataset.age_years.mean()"
]
},
{
"cell_type": "code",
"execution_count": 66,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"from pymc import Bernoulli, Normal, Uniform, invlogit, logit, deterministic, MCMC, Matplot, MAP\n",
"from pymc import Lambda, Dirichlet, Categorical, Beta, NegativeBinomial, Exponential, T\n",
"from numpy.ma import masked_values\n",
"\n",
"n_covs = 11\n",
"\n",
"def generate_model(dataset):\n",
" \n",
" # Extract data to local variables\n",
" age_years = dataset.age_years.values\n",
" school = dataset.school.values\n",
" score = dataset.score.values\n",
" non_english = dataset.non_english.astype(int).values\n",
" male = dataset.male.values\n",
" sib = dataset.sib.values\n",
" synd_or_disab = dataset.synd_or_disab.values\n",
" int_outside_option = dataset.int_outside_option.values\n",
" synd_or_disab = dataset.synd_or_disab.values\n",
" study_id = dataset.index.values\n",
" ident_by_3 = dataset.ident_by_3\n",
" program_by_6 = dataset.program_by_6\n",
" bilateral_ci = dataset.bilateral_ci\n",
" bimodal = dataset.bimodal\n",
" unimodal_ci = dataset.unimodal_ci\n",
" unimodal_ha = dataset.unimodal_ha\n",
" \n",
" # Transform some data\n",
" age_std = age_years - age_years.mean()\n",
" \n",
" # Imputation of family involvement\n",
" n_family_inv_cats = len(dataset.family_inv.unique())\n",
" p_family_inv = Dirichlet(\"p_family_inv\", np.ones(n_family_inv_cats), \n",
" value=[1./n_family_inv_cats]*(n_family_inv_cats-1))\n",
" family_inv_masked = masked_values(dataset.family_inv.fillna(0.5), value=0.5)\n",
" x_family_inv = Categorical('x_family_inv', p_family_inv, value=family_inv_masked, \n",
" observed=True)\n",
" \n",
" # Imputation of maternal education\n",
" p_mother_hs = Beta(\"p_mother_hs\", 1, 1, value=0.5)\n",
" mother_hs_masked = masked_values(dataset.mother_hs.values, value=np.nan)\n",
" x_mother_hs = Bernoulli('x_mother_hs', p_mother_hs, value=mother_hs_masked, observed=True)\n",
" \n",
" # Imputation of previous disability\n",
" p_synd_or_disab = Beta(\"p_synd_or_disab\", 1, 1, value=0.5)\n",
" synd_or_disab_masked = masked_values(dataset.synd_or_disab.values, value=np.nan)\n",
" x_synd_or_disab = Bernoulli('x_synd_or_disab', p_synd_or_disab, value=synd_or_disab_masked, \n",
" observed=True)\n",
" \n",
" # Imputation of siblings\n",
" n_sib_cats = len(dataset.sib.unique())\n",
" p_sib = Dirichlet(\"p_sib\", np.ones(n_sib_cats), value=[1./n_sib_cats]*(n_sib_cats-1))\n",
" sib_masked = masked_values(dataset.sib.fillna(0.5), value=0.5)\n",
" x_sib = Categorical('x_sib', p_sib, value=sib_masked, observed=True)\n",
" \n",
" \n",
" mean_score = Normal(\"mean_score\", 100, 0.001, value=80)\n",
" \n",
" # Indices to school random effects\n",
" unique_schools = np.unique(school)\n",
" school_index = [list(unique_schools).index(s) for s in school]\n",
"\n",
" # School random effect\n",
" sigma_school = Uniform(\"sigma_school\", 0, 1000, value=10)\n",
" tau_school = sigma_school**-2\n",
" alpha_school = Normal(\"alpha_school\", mu=0, tau=tau_school, \n",
" value=np.zeros(len(unique_schools)))\n",
" \n",
" @deterministic\n",
" def intercept(a0=mean_score, a1=alpha_school):\n",
" \"\"\"Random intercepts\"\"\"\n",
" return a0 + a1[school_index]\n",
" \n",
" # Fixed effects\n",
" beta = Normal(\"beta\", 0, 0.001, value=[0]*n_covs)\n",
"\n",
" @deterministic\n",
" def theta(b0=intercept, b=beta, x_family_inv=x_family_inv, x_sib=x_sib, \n",
" x_mother_hs=x_mother_hs, x_synd_or_disab=x_synd_or_disab):\n",
" \n",
" # Complete predictors\n",
" X = [male, ident_by_3, program_by_6, bilateral_ci, bimodal, unimodal_ci, unimodal_ha]\n",
" # Partially-imputed predictors\n",
" X += [x_family_inv, x_sib, x_mother_hs, x_synd_or_disab]\n",
" return b0 + np.dot(b, X) #+ a[child_index]\n",
" \n",
" sigma = Uniform(\"sigma\", 0, 1000, value=10)\n",
" tau = sigma**-2\n",
" \n",
" # Data likelihood\n",
" score_like = Normal(\"score_like\", mu=theta, tau=tau, value=score, observed=True)\n",
" #score_like_pred = Normal(\"score_like_pred\", mu=theta, tau=tau, value=score)\n",
" \n",
" \n",
" return locals()"
]
},
{
"cell_type": "code",
"execution_count": 67,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"M_reclang = MCMC(generate_model(receptive_language_dataset), \n",
" db='sqlite', name='reclang', dbmode='w')"
]
},
{
"cell_type": "code",
"execution_count": 68,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"iterations = 100000\n",
"burn = 90000"
]
},
{
"cell_type": "code",
"execution_count": 69,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" [-----------------100%-----------------] 100000 of 100000 complete in 1195.6 sec"
]
}
],
"source": [
"M_reclang.sample(iterations, burn)"
]
},
{
"cell_type": "code",
"execution_count": 70,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"labels = ['Male', 'Identified by 3 mo.', \n",
" 'In Program by 6 mo.', \n",
" 'Bilateral CI', 'Bimodal', 'Unimodal CI', 'Unimodal HA',\n",
" 'Reduced Family Involvement', \n",
" 'Sibling Count', 'Non-profound Hearing Loss',\n",
" 'Mother with HS Diploma','Additional Disability']\n",
"# criterion_labels[criterion]]"
]
},
{
"cell_type": "code",
"execution_count": 71,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqTgxmZlZqTgxmZlZqTgxmZlZqfx/1Av/f5gZmEYAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11a6933c8>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot([M_reclang.beta], rhat=False,\n",
" custom_labels=labels, x_range=(-15,15), main='Receptive Language')"
]
},
{
"cell_type": "code",
"execution_count": 73,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"beta:\n",
" \n",
"\tMean SD MC Error 95% HPD interval\n",
"\t------------------------------------------------------------------\n",
"\t-0.014 1.153 0.077 [-2.202 2.113]\n",
"\t4.213 1.211 0.087 [ 1.945 6.66 ]\n",
"\t7.079 1.719 0.141 [ 3.793 10.512]\n",
"\t-8.003 1.359 0.109 [-10.81 -5.494]\n",
"\t-6.393 2.042 0.179 [-10.309 -2.882]\n",
"\t-14.416 3.097 0.286 [-19.959 -8.341]\n",
"\t5.658 2.812 0.26 [ 0.512 11.589]\n",
"\t-5.67 0.538 0.03 [-6.697 -4.584]\n",
"\t-0.513 0.613 0.033 [-1.719 0.659]\n",
"\t5.077 1.368 0.109 [ 2.226 7.522]\n",
"\t-6.634 1.231 0.083 [-9.046 -4.23 ]\n",
"\t\n",
"\t\n",
"\tPosterior quantiles:\n",
"\t\n",
"\t2.5 25 50 75 97.5\n",
"\t |---------------|===============|===============|---------------|\n",
"\t-2.172 -0.861 -0.023 0.836 2.18\n",
"\t2.023 3.355 4.142 5.015 6.779\n",
"\t3.7 5.866 7.092 8.216 10.464\n",
"\t-10.652 -8.854 -7.942 -7.134 -5.244\n",
"\t-9.999 -7.739 -6.349 -5.291 -2.133\n",
"\t-20.392 -16.583 -14.364 -12.336 -8.494\n",
"\t0.461 3.575 5.635 7.481 11.589\n",
"\t-6.756 -6.031 -5.669 -5.296 -4.631\n",
"\t-1.702 -0.914 -0.513 -0.102 0.726\n",
"\t2.454 4.112 5.03 5.943 7.884\n",
"\t-9.211 -7.44 -6.597 -5.8 -4.343\n",
"\t\n"
]
}
],
"source": [
"M_reclang.beta.summary()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Expressive Language Model"
]
},
{
"cell_type": "code",
"execution_count": 74,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"7 initial_assessment_arm_1 2005-2006 0 1 0 0 3 \n",
"251 initial_assessment_arm_1 2012-2013 0 1 0 0 2 \n",
"273 initial_assessment_arm_1 2014-2015 0 0 2 0 2 \n",
"294 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"303 initial_assessment_arm_1 2011-2012 0 1 0 0 2 \n",
"\n",
" _mother_ed father_ed premature_age ... academic_year_start \\\n",
"7 2 2 8 ... 2005 \n",
"251 4 6 7 ... 2012 \n",
"273 6 6 8 ... 2014 \n",
"294 5 6 7 ... 2014 \n",
"303 6 6 8 ... 2011 \n",
"\n",
" deg_hl_below6 int_outside_option mother_hs ident_by_3 program_by_6 \\\n",
"7 0 1 0 False False \n",
"251 1 1 1 False False \n",
"273 1 1 NaN False False \n",
"294 1 1 1 False False \n",
"303 0 1 NaN False False \n",
"\n",
" jcih age_years unimodal_ci unimodal_ha \n",
"7 0 4.500000 False False \n",
"251 0 4.666667 True False \n",
"273 0 3.583333 False False \n",
"294 0 4.083333 False False \n",
"303 0 4.416667 False False \n",
"\n",
"[5 rows x 83 columns]"
]
},
"execution_count": 74,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"expressive_language_dataset = technology_subset[(technology_subset.domain=='Language') & (technology_subset.test_type=='expressive') &\n",
" (technology_subset.non_english==False)]\n",
"\n",
"expressive_language_dataset.head()"
]
},
{
"cell_type": "code",
"execution_count": 75,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mean_score = expressive_language_dataset.groupby('study_id')['score'].mean()\n",
"expressive_language_dataset = expressive_language_dataset.drop_duplicates('study_id').set_index('study_id')\n",
"expressive_language_dataset['mean_score'] = mean_score"
]
},
{
"cell_type": "code",
"execution_count": 76,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"age_years 0.00\n",
"school 0.00\n",
"score 0.00\n",
"male 0.00\n",
"time 0.04\n",
"family_inv 0.16\n",
"sib 0.06\n",
"deg_hl_below6 0.01\n",
"mother_hs 0.34\n",
"synd_or_disab 0.15\n",
"jcih 0.20\n",
"dtype: float64"
]
},
"execution_count": 76,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"(expressive_language_dataset[covs].isnull().sum() / expressive_language_dataset.shape[0]).round(2)"
]
},
{
"cell_type": "code",
"execution_count": 77,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"expressive_language_dataset = expressive_language_dataset[expressive_language_dataset.age.notnull() \n",
" & expressive_language_dataset.int_outside_option.notnull()]"
]
},
{
"cell_type": "code",
"execution_count": 78,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAF/CAYAAADn6NV5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9v8BM2122F/VMtYAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11aa5c6a0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"expressive_language_dataset.score.hist(grid=False, bins=25, figsize=(10, 6))\n",
"plt.xlabel('Standard Scores'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 79,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"M_explang = MCMC(generate_model(expressive_language_dataset), db='sqlite', name='explang', dbmode='w')"
]
},
{
"cell_type": "code",
"execution_count": 80,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" [-----------------100%-----------------] 100000 of 100000 complete in 1161.6 sec"
]
}
],
"source": [
"M_explang.sample(iterations, burn)"
]
},
{
"cell_type": "code",
"execution_count": 81,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"beta:\n",
" \n",
"\tMean SD MC Error 95% HPD interval\n",
"\t------------------------------------------------------------------\n",
"\t-0.608 1.162 0.079 [-2.866 1.539]\n",
"\t6.014 1.234 0.094 [ 3.551 8.321]\n",
"\t7.239 1.917 0.161 [ 2.9 10.575]\n",
"\t-10.202 1.172 0.081 [-12.343 -7.766]\n",
"\t-7.481 1.778 0.148 [-10.897 -4.018]\n",
"\t-15.643 2.393 0.214 [-20.17 -11.248]\n",
"\t4.27 2.725 0.252 [-1.022 9.669]\n",
"\t-4.916 0.596 0.035 [-6.123 -3.857]\n",
"\t-0.634 0.643 0.036 [-1.86 0.603]\n",
"\t6.517 1.323 0.102 [ 4.018 9.17 ]\n",
"\t-6.32 1.251 0.088 [-8.696 -3.707]\n",
"\t\n",
"\t\n",
"\tPosterior quantiles:\n",
"\t\n",
"\t2.5 25 50 75 97.5\n",
"\t |---------------|===============|===============|---------------|\n",
"\t-2.76 -1.45 -0.615 0.184 1.713\n",
"\t3.51 5.197 6.054 6.853 8.295\n",
"\t3.238 5.981 7.282 8.577 10.96\n",
"\t-12.577 -11.039 -10.232 -9.396 -7.899\n",
"\t-10.895 -8.753 -7.463 -6.288 -3.999\n",
"\t-20.448 -17.373 -15.479 -13.926 -11.352\n",
"\t-0.93 2.321 4.061 6.167 9.799\n",
"\t-6.09 -5.328 -4.932 -4.505 -3.77\n",
"\t-1.845 -1.078 -0.645 -0.184 0.652\n",
"\t4.082 5.683 6.403 7.402 9.273\n",
"\t-8.829 -7.166 -6.327 -5.518 -3.79\n",
"\t\n"
]
}
],
"source": [
"M_explang.beta.summary()"
]
},
{
"cell_type": "code",
"execution_count": 82,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11aad97f0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot([M_explang.beta], custom_labels=labels, rhat=False, main='Expressive Language', x_range=(-15,15))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Articulation Model"
]
},
{
"cell_type": "code",
"execution_count": 84,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"4 initial_assessment_arm_1 2005-2006 0 1 0 0 3 \n",
"269 initial_assessment_arm_1 2014-2015 0 0 2 0 2 \n",
"290 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"371 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"504 initial_assessment_arm_1 2014-2015 0 1 0 0 0 \n",
"\n",
" _mother_ed father_ed premature_age ... academic_year_start \\\n",
"4 2 2 8 ... 2005 \n",
"269 6 6 8 ... 2014 \n",
"290 5 6 7 ... 2014 \n",
"371 6 6 8 ... 2014 \n",
"504 4 4 8 ... 2014 \n",
"\n",
" deg_hl_below6 int_outside_option mother_hs ident_by_3 program_by_6 \\\n",
"4 0 1 0 False False \n",
"269 1 1 NaN False False \n",
"290 1 1 1 False False \n",
"371 0 1 NaN True False \n",
"504 1 0 1 False False \n",
"\n",
" jcih age_years unimodal_ci unimodal_ha \n",
"4 0 4.500000 False False \n",
"269 0 3.583333 False False \n",
"290 0 4.083333 False False \n",
"371 0 3.916667 False False \n",
"504 0 4.500000 False False \n",
"\n",
"[5 rows x 83 columns]"
]
},
"execution_count": 84,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"articulation_dataset = technology_subset[(technology_subset.domain=='Articulation') & \n",
" (technology_subset.non_english==False)]\n",
"articulation_dataset.head()"
]
},
{
"cell_type": "code",
"execution_count": 85,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mean_score = articulation_dataset.groupby('study_id')['score'].mean()\n",
"articulation_dataset = articulation_dataset.drop_duplicates('study_id').set_index('study_id')\n",
"articulation_dataset['mean_score'] = mean_score"
]
},
{
"cell_type": "code",
"execution_count": 86,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"age_years 0.00\n",
"school 0.00\n",
"score 0.00\n",
"male 0.00\n",
"time 0.04\n",
"family_inv 0.17\n",
"sib 0.05\n",
"deg_hl_below6 0.01\n",
"mother_hs 0.32\n",
"synd_or_disab 0.14\n",
"jcih 0.21\n",
"dtype: float64"
]
},
"execution_count": 86,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"(articulation_dataset[covs].isnull().sum() / articulation_dataset.shape[0]).round(2)"
]
},
{
"cell_type": "code",
"execution_count": 87,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"articulation_dataset = articulation_dataset[articulation_dataset.age.notnull()\n",
" & articulation_dataset.int_outside_option.notnull()]"
]
},
{
"cell_type": "code",
"execution_count": 88,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAF/CAYAAADn6NV5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGt1JREFUeJzt3X+w5WddH/D3ZzfZ8GtDJGqCihC4kdqZtggzMGAQk93i\nkCpKFWGmKP4oxSDg1GkqP+yETSkVKBkdGWPHKhWrKGiB0QnsZrMJu4QILWjiSAkuEyepMeuwgfWi\nu1k3+/SP8116s9y799zN+d7n3pzXa+bMPec59zzfz33m7j3vfZ7veb7VWgsAAP1s6V0AAMC8E8gA\nADoTyAAAOhPIAAA6E8gAADoTyAAAOhs1kFXVj1fVJ6vq41V1+dC2s6oOVNX+qrpizOMDAGwGNeY+\nZFV1R5JnJHlcko8m+c4kH0+yI0kl2d1a+67RCgAA2ATOGbn/O5LsTPKNSXYnuTTJna21Y0lSVQer\naqG1dnDkOgAANqyxA9mBJK/MZGn0fUkuTHKkqq7LZIbsyNAmkAEAc2u0QFZVT0tyRWvtZcPjm5O8\nPskFSa7KJJBdn+TwSn3cdNNNrusEAGwaO3bsqLN53ZgzZFuSPD5JqurcTILYwUyWLZNJIFt1ufKZ\nz3zmiCUCAMzGZz7zmbN+7WiBrLX2F8MnKW/LJHz9YmvtaFVdm2RvkpZk11jHBwDYLEY9h6y19rYk\nbzutbU+SPWMeFwBgM7ExLABAZwIZAEBnAhkAQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBnAhkA\nQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBn\nAhkAQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBnAhkAQGcCGQBAZwIZAEBn5/QuAABWct/iAzm0\neHwmfV20fVsu3n7eTPqCWRPIANiwDi0ez9U3HJxJX++8ckEgY8OyZAkA0JlABgDQmUAGANDZaIGs\nqs6vqpurat/w9ctD+86qOlBV+6vqirGODwCwWYx2Un9r7W+TXJ4kVfVPk7yuqirJriQ7klSS3Un2\njVUDAMBmsF5Llq9L8stJLk1yZ2vtWGvtaJKDVbWwTjUAAGxIo297UVVPSPKk1todVfXcJEeq6rpM\nZsiOJLkwyWw+0wwAsAmtxz5k/ybJrw33Dye5IMlVmQSy64c2gEe0WW5wmtjkFB5pRg1kVbU1yfcl\nef7QdDCTZctkEsgWWmtmx4BHvFlucJrY5BQeacaeIXtJkj9srZ1MktbayaralWRvkpbJCf4AAHNt\n1EDWWvv9ZdpuTHLjmMcFANhMbAwLANCZQAYA0JlABgDQmUAGANDZeuxDBsCMbdtauf3exZn0ZU+z\ntbOvHLMmkAFsQvcfPZFde++aSV/2NFs7+8oxa5YsAQA6E8gAADqzZAnAzMz63KrjD56cWV+wkQlk\nAMzMrM+tumbnJTPrCzYyS5YAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlk\nAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAA\nnQlkAACdCWQAAJ0JZAAAnQlkAACdjRrIquqbq2pfVe2vqncNbTur6sDQdsWYxwcA2AzOGbn//5Lk\nza2125KkqirJriQ7klSS3Un2jVwDAMCGNtoMWVVtSbJwKowNLk1yZ2vtWGvtaJKDVbUwVg0AAJvB\nmDNk35DkUVX1wSTnJ3l3kvuSHKmq6zKZITuS5MIkB0esAwBgQxszkB1O8uUkPzgc59YkP5HkgiRX\nZRLIrh++DwBgbo22ZNlaO5HkniRPbK0dT3Isk5mwS4dvqUyWNM2OAQBzbeyT+t+Q5Neq6vwk72+t\nHa2qa5PsTdIyOcEfAGCujRrIWmt3J7nytLY9SfaMeVwAgM3ExrAAAJ0JZAAAnQlkAACdCWQAAJ0J\nZAAAnQlkAACdCWQAAJ0JZAAAnY29Uz/ApnXf4gM5tHh8Jn0df/DkTPoZw7atldvvXZxJXxv554SN\nTCADWMGhxeO5+obZXG73mp2XzKSfMdx/9ER27b1rJn1t5J8TNjJLlgAAnQlkAACdCWQAAJ0JZAAA\nnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0JZAAAnQlkAACdCWQAAJ0J\nZAAAnQlkAACdCWQAAJ0JZAAAnZ3TuwCA+xYfyKHF4zPp66Lt23Lx9vNm0hfAehHIgO4OLR7P1Tcc\nnElf77xyQSADNh1LlgAAnQlkAACdWbIEYC5s21q5/d7FmfR1/MGTM+kHThk1kFXVe5L8oyRHk/z3\n1tp7q2pnkmuStCRvaa3tG7MGAEiS+4+eyK69d82kr2t2XjKTfuCUsWfIWpIfbq3dkyRVVUl2JdmR\npJLsTiKQAQBzbexzyOq0Y1ya5M7W2rHW2tEkB6tqYeQaAAA2tLFnyBaT/E5VHU7ys0kuTHKkqq7L\nJKwdGdpm83l3AIBNaNRA1lp7fZJU1TOSvDPJv09yQZKrMglk1yc5PGYNAAAb3Xpte3EsyT8k+UIm\ny5bJJJAttNbMjgEAc23sT1n+bpInZrJ0+dOttZNVtSvJ3kxO+N815vEBADaDsZcsX75M241Jbhzz\nuAAAm4md+gEAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOls1kFXV49ejEACAeTXNDNnt\nVfUbVfWc0asBAJhD0wSyhSR/mOTNVfW/q+qnqupxI9cFADA3Vg1krbUTrbUPttZenOQ1Sd6Q5O6q\nepdgBgDw8E1zDtm2qnpZVX00yS8neVuSJye5NckfjFwfAMAj3jQXFz+Y5CNJ3txa+/SS9v9ZVa8d\npywAgPkxTSD7x621r6zwnEAGAPAwTXMO2UphLK21z862HACA+TPNOWQ/uUybmTEAgBmZZtuLH1+m\n7aWzLgQAYF6d7aWTts60CgCAOTZNIDtUVS889aCqXpLki+OVBAAwX6b5lOW/TfLhqnpLJjNj5yX5\n/jGLAgCYJ6sGstba3VX1rCRPH5rubK2dHLcsAID5Mc0MWYYA9n9GrgUAYC6tGsiq6glJfiDJBUvb\nW2vXjVUUAMA8mWaGbHeSP09y18i1AADMpWkC2Vdaaz82diEAAPNqmm0vPl1VT1/92wAAOBvTzJD9\nsyQ3VtWfLm1srb14nJIAAObLNIHsraNXAQAwx6bZh+xj61EIAMC8mupallX1lKp60ZLHjx2vJACA\n+bJqIKuqVyR5X5L/vKT5I6NVBAAwZ6aZIXtNkhck+dKSthqnHACA+TNNIDvRWjt+6kFVPS7Jo8cr\nCQBgvkzzKcs/rqpfSHJ+VX1fkquT/Pa4ZQFJct/iAzm0eHz1b5zCRdu35eLt582kr41s29bK7fcu\nzqSv4w+enEk/AKuZJpC9Icmrkvxlklckub619r5pD1BV25J8Psk7Wmu/UlU7k1yTpCV5S2tt35qr\nhjlxaPF4rr7h4Ez6eueVC3MRyO4/eiK79s7mSm/X7LxkJv0ArGaabS9OJvmvw+1s/FSSTydJVVWS\nXUl2ZHIe2u4kAhkAMNem2vbibFXVo5O8MMmHh6ZLk9zZWjvWWjua5GBVLYxZAwDARrfqDFlVLWay\nvPgQrbXzp+j/9UneneSi4fGFSY5U1XWZzJAdGdpmsyYDALAJTbNkuX3p46p6XpJnrfa6qjo/yfNb\na2+vqldmEsAOJ7kgyVXD4+uHNgCAuTXNSf0P0Vr7RFW9fIpvvSzJeVX1O0memmRrkgOZLFsmk0C2\n0FozOwYAzLVpliyfeVrTNyZ5zmqva63dkOSGoY8fTfK41todVXVtkr2ZLIPuWnPFAACPMNPMkL3r\ntMf3J3njWg7SWnvvkvt7kuxZy+sBAB7JpjmH7PL1KAQAYF6Nuu0FAACrm+YcspuzzLYXp7TWrphp\nRQAAc2aac8g+leTL+f/nfb10+PqBUSoCAJgz0wSy72itvXDJ409X1S2ttTeMVRQAwDyZ5hyyb6mq\nbzj1oKoen8nu+gAAzMA0M2RvT/KnVXVTJueSPT/JNaNWBQAwR6bZ9uI3q2pPkmdnEsiubq39zeiV\nAQDMiakundRa++skHx65FgCAuTTVPmRV9Yqq2jXcr+EC4wAAzMCqgayq3pXJtSu/J0laay3JO0au\nCwBgbkwzQ/bs1trrkhxd0rbiRrEAAKzNNIFsS1WdkyGEVdXTMuW5ZwAArG6aYPUrSfYmedKwfPnS\nJP961KoAAObINNte/HZV/UmSHUlOJHlBa+2u0SsDAJgT02578dkknx25FgCAuTTNpyyftB6FAADM\nq2lO6v/D0asAAJhj0wSyY6NXAQAwx6YJZP+tqt5VVU9Yehu9MgCAOTHNSf1vHr7+yyVtLclTZ18O\nAMD8mWbbi0vWoxAAgHk11cXFAQAYz4qBrKp+b8n9161POQAA8+dMM2TftOT+S8YuBABgXp0pkLV1\nqwIAYI6d6aT+b6mqn01SSb51uP9VrbXrRq0MAGBOnCmQ/WaS7cP931pyHwCAGVoxkLXWdq1nIQAA\n82qajWEBHuK+xQdyaPH4zPo7/uDJmfUFsBkJZMCaHVo8nqtvODiz/q7Zaf9pYL7ZGBYAoDOBDACg\ns1EDWVW9tar2VdWNVXXJ0Lajqg5U1f6qumLM4wMAbAajnkPWWvv5JKmq70zyc1V1VZJrk+zIZH+z\n3Un2jVkDAMBGt15Lls9J8rkklya5s7V2rLV2NMnBqlpYpxoAADak0T9lWVUfS/LEJJcleVqSI1V1\nXSYzZEeSXJhkdh/XAgDYZEafIWutvSDJv0ry3iRfTHJBkjcNt69LcnjsGgAANrL1WrI8lMnFyr+Q\nybJlMpkhW2itmR0DAObaqEuWVfV7Sb4+ydEkr22tnayqXUn2ZhLQXJ4JAJh7Y3/K8mXLtN2Y5MYx\njwsAsJm4dBLMiW1bK7ffuziTvlx7EmC2BDKYE/cfPZFde++aSV+uPQkwWy6dBADQmUAGANCZQAYA\n0JlzyJh79y0+kEOLx2fS10Xbt+Xi7efNpC8A5odAxtw7tHg8V98wm/2J33nlgkAGwJpZsgQA6Ewg\nAwDozJIlAHQ2y42bncu6OQlkANDZLDdudi7r5mTJEgCgM4EMAKAzgQwAoDOBDACgM4EMAKAzgQwA\noDOBDACgM4EMAKAzgQwAoDOBDACgM4EMAKAzgQwAoDOBDACgM4EMAKAzgQwAoLNzehcAjyTbtlZu\nv3dxZv0df/DkzPoCYOMSyGCG7j96Irv23jWz/q7ZecnM+gJg47JkCQDQmUAGANCZQAYA0JlABgDQ\nmUAGANCZQAYA0NmogayqfrWqbq6qW6rqkqFtR1UdqKr9VXXFmMcHANgMRt2HrLX2U0lSVZcnubqq\nfjrJtUl2JKkku5PsG7MGAICNbr2WLBeTHE9yaZI7W2vHWmtHkxysqoV1qgEAYENar536fzLJLyW5\nMMmRqroukxmyI0PbwXWqAwBgwxk9kFXV92YyK/a5qvq2JBckuSqTQHZ9ksNj1wAAsJGNfVL/s5J8\nd2vtF4emg5ksWyaTQLbQWjM7BgDMtbFnyD6Q5J6qujnJHa21n6mqa5PsTdKS7Br5+AAAG97Yn7J8\n6jJte5LsGfO4AACbiY1hAQA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAA\nOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoT\nyAAAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOhPIAAA6E8gAADoTyAAAOjundwEAwOxs\n21q5/d7FmfR10fZtuXj7eTPpizMTyADgEeT+oyeya+9dM+nrnVcuCGTrxJIlAEBnAhkAQGejBrKq\nuqyqPlVV71jStqOqDlTV/qq6YszjAwBsBmOfQ3ZekrcleV6SVFUluTbJjiSVZHeSfSPXAACwoY06\nQ9ZauynJl5Y0XZrkztbasdba0SQHq2phzBoAADa69f6U5YVJjlTVdZnMkB0Z2g6ucx0AABvGegey\nw0kuSHJVJoHs+qENAGBurdenLGv4ejCTZctTbQutNbNjAMBcG3WGrKp+LsmLklxUVee31l5dVdcm\n2ZukJdk15vEBADaDUQNZa+3tSd5+WtueJHvGPC4AwGZiY1gAgM4EMgCAzgQyAIDOBDIAgM4EMgCA\nzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIAgM4E\nMgCAzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIAgM4EMgCAzgQyAIDOBDIA\ngM4EMgCAzs7pXcBqPnXP3z7sPs7dUnn6Nz4mjzl36wwqAgCYrQ0fyH5+9xcedh8Xb9+Wd3//05Nz\nZ1AQAMCMWbIEAOhMIAMA6EwgAwDorEsgq6odVXWgqvZX1RU9agAA2CjW/aT+qqok1ybZkaSS7E6y\nb73rAADYKHrMkF2a5M7W2rHW2tEkB6tqoUMdAAAbQo9tLy5McqSqrstkhuzI0HZwuW9+9XO++WEf\n8LHbtmRLPexuAABGUa219T1g1bcleWOSqzIJZNcneWtr7WsC2U033bS+xQEAPAw7duw4qymgHoFs\nS5L9SXZmsmS6p7V22boWAQCwgaz7kmVr7WRV7UqyN0lLsmu9awAA2EjWfYYMAICHsjEsAEBnAhkA\nQGcCGQBAZxs2kLm80vSq6ler6uaquqWqLhnajN8aVdW2qvrLqnrN8HinMZxOVX1zVe0bxupdQ5vx\nm0JV/XhVfbKqPl5Vlw9txm4FVXVZVX2qqt6xpG3Zv3f+Dn6tFcbva95Dhnbjd5rlxm9of8j7x9C2\ntvFrrW24Wyb7k92a5FFJHp1kf++aNsMtyeVJfsX4nfX4vT7JHyR5jTFc89i9L8lzlzw2ftOP3R2Z\n/Of4/CSfMHarjteOJD+Q5B3D42XHyzhON36nPXd5kuuN39rHb+n7x9mO30adIXN5pbOzmOR4jN+a\nVdWjk7wwyYeHJmM4pWFvwYXW2m1Lmo3f9O7IZF/GF2dybV9jdwattZuSfGlJ00rjZRyXscz4LbWY\n5IHhvvFbxnLjt8z7R3IW49fj0knTWNPllfiqn0zySzF+Z+P1Sd6d5KLhsTGc3jckeVRVfTCTWZ53\nJ7kvxm9aB5K8MpNZsvfF795arTReW1ZoN44rO/Uekvg9XIvT3z+Ssxi/jRrIDie5IA+9vNLhrhVt\ncFX1vZmk8c8Nl6cyflOqqvOTPL+19vaqemUmY+Z3cHqHk3w5yQ9m8jfl1iQ/EeO3qqp6WpIrWmsv\nGx7fnMkfd2M3vZX+rW5ZoZ1lLH0PGZr8DZzCMu8fp6x5/DZqIDuYyXRfMvlBFtoy17pkoqqeleS7\nW2v/bmgyfmtzWZLzqup3kjw1ydZMZi2M4RRaayeq6p4kT2yt/VVVHYvfwWltSfL4JKmqczP5A27s\npnPqeoHLjtewlG4cV/bV6y0u8x6S+D1czanx+5r3j6q6Jcnnssbx25CBrLm80lp9IMk9w/+u72it\n/UxVXRvjN5XW2g1JbkiSqvrRJI9rrd1hDNfkDUl+bfjf4vtba0eN3+paa38xfALrtkz+aP+isTuz\nqvq5JC9KclFVnd9ae/Vy4+V9ZHnLjV+WeQ8xfstbYfxOf//47PB4TePn0kkAAJ1t1E9ZAgDMDYEM\nAKAzgQwAoDOBDACgM4EMAKAzgQw4o6p6TFW9t6puGy6U+9rTnn98VV01cg2LI/T5rOFj/tN+/xuH\niwrvr6rfmHU9wHzbkPuQARvK1Unubq396ArPf10mF2S/fsQaxtqfZ6p+q+oFSb6ntfbskeoA5pwZ\nMmA1lckO8l/7RNVzk7w/yVOGmaMPnvb8W6pqd1V9pqr+qKrOW/LcYlW9qqo+WlWfH/o69dwzq+rT\nVXVzVb01D91VfHtVvaeq9lTV56rqP512zPdU1Zuq6mPDjNYPLXnulVX151W1J8kPr3EMHl1Vj11h\nHF4w1Hqgqm6tqmcseW7H0HZgGINvXfLck6vqz6rq2qr6ZFXtXfLclqp6xzCun6iqV5x2zJ8dXrO/\nqj6yhp8F2Ihaa25ubm4r3pI8NslvJfl0kh9Z5vknZ7K793Kv/fol9z+U5OVLHv9Dku8f7v9Ykvcu\nee7PMrmUSzK5NMnx0/p9wvD10Un+KpPLNp167j1Jbs5kx+ylr/mmJP/3VE1J3phk3xrG4U1J7hy+\nPva0n/8vkjxpmddcmOSuU/Ul+YEk+0977bEkP7TMa1+d5BeG+9uS3JbkKcPjxyf5myTn9P79cHNz\nm83NDBlwRq21v2ut/UiSlyT5vqr69TW8/EvD7NGrkjwmyROXPHe0tfbh4f5dGWbhquqCJOe31m4Z\njv/xTELLUv9QVf8ik4uYH0ty8WnP/3Jr7SuntT07yU2ttS8Oj/es4edIa+1tSZ6V5Nwkt1bVo4an\nrkzygdbaPcu87LlJPt5a++uhjw8leeppM22fb639/jKvfWGSfz6c57Y7k/D57UM/RzK5XMsNVfXa\nqvr6tfwswMYjkAFTaa3dneTlSV5cVauef1pVj0nyx0m+K8kXMrlYcZ3xRRMPrtLvP0ny8SRPSvKn\nSb64hn6n+b4Vtda+0lrbleTuJM9b8tSZxuP0v7OV6c5dO5HkLa21y4fbM1prX12abK39WJIfGb7v\nk1X1lCn6BDYogQw4oyFYnfLtSQ611k4saTuW5MKq2jJ8/6nQ8/RMlhr/Y5LPJPmOPDQQLRuOWmuL\nSQ5V1XcO/X1vJrNrp+xM8kettV9N8rdJLlmpr9PcluSyYQYuSV46xWsy1HBOVZ073N+e5CmZBMwk\n+UiSH66qhWVe+okkz6uqbxle+9JMZsT+fmn3Kxz2Q0murqrHrVDTltbaoWEcPp9h9gzYnHzKEljN\ni6vq6iRfSfL3OS3ItNYOVdXHkvxJVd2X5OeT/K8ktye5u6puz2RG6ZY8dGnxTLNEr0ry61V1bHjd\n0gDzu0k+VFWXJ/lckv3T9Nta+2JV/YckB6rqcCazd9N6WpL/UVVHh8dvGmYM01r7y6p6ZZLfGMLo\nyeH5W1tr91fVTyR5f1WdTHIkyemfVl2p3vdV1cVJbhmO25K8qLX2d8NxbqqqrUkeleRjST66hp8H\n2GCqtbE+TQ4AwDQsWQIAdCaQAQB0JpABAHQmkAEAdCaQAQB0JpABAHQmkAEAdCaQAQB09v8AtzJ3\nV0DROUAAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10b4b1c88>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"articulation_dataset.score.hist(grid=False, bins=25, figsize=(10, 6))\n",
"plt.xlabel('Standard Scores'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 89,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"M_articulation = MCMC(generate_model(articulation_dataset), db='sqlite', name='articulation', dbmode='w')"
]
},
{
"cell_type": "code",
"execution_count": 90,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" [-----------------100%-----------------] 100000 of 100000 complete in 1115.2 sec"
]
}
],
"source": [
"M_articulation.sample(iterations, burn)"
]
},
{
"cell_type": "code",
"execution_count": 91,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11a6d9fd0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot([M_articulation.beta], \n",
" custom_labels=labels, rhat=False, main='Articulation', x_range=(-15,15))"
]
},
{
"cell_type": "code",
"execution_count": 93,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"beta:\n",
" \n",
"\tMean SD MC Error 95% HPD interval\n",
"\t------------------------------------------------------------------\n",
"\t1.025 1.393 0.099 [-1.822 3.482]\n",
"\t6.268 1.601 0.125 [ 3.039 9.445]\n",
"\t4.507 2.406 0.211 [-0.181 9.25 ]\n",
"\t-5.6 1.654 0.132 [-8.819 -2.345]\n",
"\t-3.846 2.327 0.205 [-8.176 0.582]\n",
"\t-7.661 3.711 0.349 [-15.473 -1.608]\n",
"\t3.64 2.489 0.217 [-1.082 9.486]\n",
"\t-3.331 0.614 0.034 [-4.58 -2.136]\n",
"\t0.007 0.714 0.039 [-1.352 1.393]\n",
"\t5.354 1.584 0.124 [ 2.196 8.587]\n",
"\t-7.866 1.433 0.103 [-10.788 -5.272]\n",
"\t\n",
"\t\n",
"\tPosterior quantiles:\n",
"\t\n",
"\t2.5 25 50 75 97.5\n",
"\t |---------------|===============|===============|---------------|\n",
"\t-1.798 0.111 1.118 1.982 3.537\n",
"\t2.927 5.227 6.344 7.344 9.399\n",
"\t-0.165 2.872 4.555 6.079 9.415\n",
"\t-8.918 -6.669 -5.574 -4.553 -2.411\n",
"\t-8.119 -5.561 -3.958 -2.179 0.696\n",
"\t-15.325 -10.014 -7.205 -4.965 -1.325\n",
"\t-1.852 2.085 3.707 5.221 8.908\n",
"\t-4.549 -3.748 -3.339 -2.93 -2.1\n",
"\t-1.391 -0.483 0.023 0.495 1.365\n",
"\t2.12 4.372 5.339 6.402 8.558\n",
"\t-10.712 -8.899 -7.858 -6.919 -5.118\n",
"\t\n"
]
}
],
"source": [
"M_articulation.beta.summary()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Expressive Vocabulary Model"
]
},
{
"cell_type": "code",
"execution_count": 94,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"5 initial_assessment_arm_1 2005-2006 0 1 0 0 3 \n",
"19 initial_assessment_arm_1 2007-2008 0 0 0 0 3 \n",
"246 initial_assessment_arm_1 2012-2013 0 0 0 0 2 \n",
"270 initial_assessment_arm_1 2014-2015 0 0 2 0 2 \n",
"291 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"\n",
" _mother_ed father_ed premature_age ... academic_year_start \\\n",
"5 2 2 8 ... 2005 \n",
"19 4 4 8 ... 2007 \n",
"246 6 6 8 ... 2012 \n",
"270 6 6 8 ... 2014 \n",
"291 5 6 7 ... 2014 \n",
"\n",
" deg_hl_below6 int_outside_option mother_hs ident_by_3 program_by_6 \\\n",
"5 0 1 0 False False \n",
"19 0 1 1 False False \n",
"246 1 1 NaN False False \n",
"270 1 1 NaN False False \n",
"291 1 1 1 False False \n",
"\n",
" jcih age_years unimodal_ci unimodal_ha \n",
"5 0 4.500000 False False \n",
"19 0 4.583333 False False \n",
"246 0 4.000000 False False \n",
"270 0 3.583333 False False \n",
"291 0 4.083333 False False \n",
"\n",
"[5 rows x 83 columns]"
]
},
"execution_count": 94,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"expressive_vocab_dataset = technology_subset[(technology_subset.domain=='Expressive Vocabulary') & \n",
" (technology_subset.non_english==False)]\n",
"expressive_vocab_dataset.head()"
]
},
{
"cell_type": "code",
"execution_count": 95,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mean_score = expressive_vocab_dataset.groupby('study_id')['score'].mean()\n",
"expressive_vocab_dataset = expressive_vocab_dataset.drop_duplicates('study_id').set_index('study_id')\n",
"expressive_vocab_dataset['mean_score'] = mean_score"
]
},
{
"cell_type": "code",
"execution_count": 96,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"age_years 4\n",
"school 0\n",
"score 0\n",
"male 0\n",
"time 40\n",
"family_inv 153\n",
"sib 49\n",
"deg_hl_below6 8\n",
"mother_hs 312\n",
"synd_or_disab 124\n",
"jcih 200\n",
"dtype: int64"
]
},
"execution_count": 96,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"expressive_vocab_dataset[covs].isnull().sum()"
]
},
{
"cell_type": "code",
"execution_count": 97,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"expressive_vocab_dataset = expressive_vocab_dataset[expressive_vocab_dataset.age.notnull() \n",
" & expressive_vocab_dataset.male.notnull()\n",
" & expressive_vocab_dataset.int_outside_option.notnull()]"
]
},
{
"cell_type": "code",
"execution_count": 98,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAmkAAAF/CAYAAAASFl7JAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/X8bYAJr+cEd0wAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10ea69860>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"expressive_vocab_dataset.score.hist(grid=False, bins=25, figsize=(10, 6))\n",
"plt.xlabel('Standard Scores'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 99,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"M_expressive_vocab = MCMC(generate_model(expressive_vocab_dataset), db='sqlite', name='expressive_vocab', dbmode='w')"
]
},
{
"cell_type": "code",
"execution_count": 100,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" [-----------------100%-----------------] 100000 of 100000 complete in 1208.7 sec"
]
}
],
"source": [
"M_expressive_vocab.sample(iterations, burn)"
]
},
{
"cell_type": "code",
"execution_count": 101,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuKWpGZcTigAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11a7b9b70>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot([M_expressive_vocab.beta], \n",
" custom_labels=labels, rhat=False, main='Expressive Vocabulary', x_range=(-15,15))"
]
},
{
"cell_type": "code",
"execution_count": 102,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrFScmMzMrFScmMzMrFScmMzMrFT+f9cKJcw/PGs9AAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x1087fab70>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot([M_expressive_vocab.beta], \n",
" custom_labels=labels, rhat=False, main='Expressive Vocabulary', x_range=(-10,10))"
]
},
{
"cell_type": "code",
"execution_count": 103,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"beta:\n",
" \n",
"\tMean SD MC Error 95% HPD interval\n",
"\t------------------------------------------------------------------\n",
"\t0.975 1.16 0.08 [-1.414 3.056]\n",
"\t6.812 1.28 0.098 [ 4.052 9.213]\n",
"\t5.863 2.085 0.177 [ 1.974 10.077]\n",
"\t-5.692 1.275 0.094 [-8.147 -3.308]\n",
"\t-5.099 1.806 0.148 [-8.481 -1.372]\n",
"\t-15.035 3.036 0.282 [-21.84 -9.861]\n",
"\t6.66 2.365 0.209 [ 2.757 11.604]\n",
"\t-5.144 0.532 0.029 [-6.177 -4.129]\n",
"\t-1.395 0.684 0.045 [-2.696 -0.007]\n",
"\t6.207 1.391 0.111 [ 3.567 8.881]\n",
"\t-5.2 1.222 0.093 [-7.604 -2.894]\n",
"\t\n",
"\t\n",
"\tPosterior quantiles:\n",
"\t\n",
"\t2.5 25 50 75 97.5\n",
"\t |---------------|===============|===============|---------------|\n",
"\t-1.268 0.213 0.98 1.75 3.277\n",
"\t4.152 6.004 6.826 7.654 9.359\n",
"\t1.191 4.751 5.903 7.196 9.834\n",
"\t-8.101 -6.568 -5.659 -4.84 -3.217\n",
"\t-8.628 -6.282 -5.102 -3.921 -1.378\n",
"\t-21.495 -16.902 -14.941 -12.865 -9.248\n",
"\t2.254 5.042 6.594 8.32 11.314\n",
"\t-6.209 -5.481 -5.156 -4.793 -4.141\n",
"\t-2.719 -1.868 -1.406 -0.943 -0.013\n",
"\t3.658 5.193 6.143 7.181 9.14\n",
"\t-7.714 -6.039 -5.138 -4.33 -2.985\n",
"\t\n"
]
}
],
"source": [
"M_expressive_vocab.beta.summary()"
]
},
{
"cell_type": "code",
"execution_count": 104,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Could not calculate Gelman-Rubin statistics. Requires multiple chains of equal length.\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaIAAAEgCAYAAAD2c3e8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11aa62da0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Matplot.summary_plot(M_explang.alpha_school, custom_labels=[' ']*42, main='Expressive language school effects', \n",
" x_range=(-25, 25))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Receptive Vocabulary Model"
]
},
{
"cell_type": "code",
"execution_count": 106,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
" redcap_event_name academic_year hl male _race prim_lang sib \\\n",
"20 initial_assessment_arm_1 2007-2008 0 0 0 0 3 \n",
"82 initial_assessment_arm_1 2009-2010 0 0 0 0 2 \n",
"247 initial_assessment_arm_1 2012-2013 0 0 0 0 2 \n",
"271 initial_assessment_arm_1 2014-2015 0 0 2 0 2 \n",
"292 initial_assessment_arm_1 2014-2015 0 0 0 0 1 \n",
"\n",
" _mother_ed father_ed premature_age ... academic_year_start \\\n",
"20 4 4 8 ... 2007 \n",
"82 4 4 8 ... 2009 \n",
"247 6 6 8 ... 2012 \n",
"271 6 6 8 ... 2014 \n",
"292 5 6 7 ... 2014 \n",
"\n",
" deg_hl_below6 int_outside_option mother_hs ident_by_3 program_by_6 \\\n",
"20 0 1 1 False False \n",
"82 0 1 1 False False \n",
"247 1 1 NaN False False \n",
"271 1 1 NaN False False \n",
"292 1 1 1 False False \n",
"\n",
" jcih age_years unimodal_ci unimodal_ha \n",
"20 0 4.583333 False False \n",
"82 0 3.833333 False False \n",
"247 0 4.000000 False False \n",
"271 0 3.583333 False False \n",
"292 0 4.083333 False False \n",
"\n",
"[5 rows x 83 columns]"
]
},
"execution_count": 106,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_vocab_dataset = technology_subset[(technology_subset.domain=='Receptive Vocabulary') & \n",
" (technology_subset.non_english==False)]\n",
"receptive_vocab_dataset.head()"
]
},
{
"cell_type": "code",
"execution_count": 107,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mean_score = receptive_vocab_dataset.groupby('study_id')['score'].mean()\n",
"receptive_vocab_dataset = receptive_vocab_dataset.drop_duplicates('study_id').set_index('study_id')\n",
"receptive_vocab_dataset['mean_score'] = mean_score"
]
},
{
"cell_type": "code",
"execution_count": 108,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"age_years 4\n",
"school 0\n",
"score 0\n",
"male 0\n",
"time 37\n",
"family_inv 158\n",
"sib 49\n",
"deg_hl_below6 10\n",
"mother_hs 318\n",
"synd_or_disab 125\n",
"jcih 202\n",
"dtype: int64"
]
},
"execution_count": 108,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_vocab_dataset[covs].isnull().sum()"
]
},
{
"cell_type": "code",
"execution_count": 109,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"receptive_vocab_dataset = receptive_vocab_dataset[receptive_vocab_dataset.age.notnull() & \n",
" receptive_vocab_dataset.male.notnull() &\n",
" receptive_vocab_dataset.int_outside_option.notnull()]"
]
},
{
"cell_type": "code",
"execution_count": 110,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x108826da0>"
]
},
"execution_count": 110,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEBCAYAAABysL6vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nV0IAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x1087e0b70>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"receptive_vocab_dataset.age_int.hist(bins=40)"
]
},
{
"cell_type": "code",
"execution_count": 111,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"(922, 83)"
]
},
"execution_count": 111,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"receptive_vocab_dataset.shape"
]
},
{
"cell_type": "code",
"execution_count": 112,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAmkAAAF/CAYAAAASFl7JAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10ec22550>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"receptive_vocab_dataset.score.hist(grid=False, bins=25, figsize=(10, 6))\n",
"plt.xlabel('Standard Scores'); plt.ylabel('Frequency');"
]
},
{
"cell_type": "code",
"execution_count": 113,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"M_receptive_vocab = MCMC(generate_model(receptive_vocab_dataset), db='sqlite', name='receptive_vocab', dbmode='w')"
]
},
{
"cell_type": "code",
"execution_count": 114,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" [-----------------100%-----------------] 100000 of 100000 complete in 1169.1 sec"
]
}
],
"source": [
"M_receptive_vocab.sample(iterations, burn)"
]
},
{
"cell_type": "code",
"execution_count": 115,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAEgCAYAAAD/mNfGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x10eaac4a8>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"#order = np.argsort(M.beta.stats()['mean'])[::-1]\n",
"Matplot.summary_plot([M_receptive_vocab.beta], rhat=False,\n",
" custom_labels=labels, main='Receptive Vocabulary', x_range=(-15,15))"
]
},
{
"cell_type": "code",
"execution_count": 117,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"beta:\n",
" \n",
"\tMean SD MC Error 95% HPD interval\n",
"\t------------------------------------------------------------------\n",
"\t0.385 1.001 0.06 [-1.625 2.257]\n",
"\t4.984 1.173 0.087 [ 2.65 7.07]\n",
"\t5.367 1.732 0.145 [ 1.841 8.586]\n",
"\t-5.109 1.221 0.095 [-7.366 -2.551]\n",
"\t-6.193 1.888 0.166 [-9.57 -2.332]\n",
"\t-14.416 2.514 0.232 [-18.62 -9.371]\n",
"\t6.853 2.164 0.194 [ 2.9 11.565]\n",
"\t-4.353 0.525 0.031 [-5.261 -3.238]\n",
"\t-1.132 0.564 0.031 [-2.225 -0.051]\n",
"\t5.11 1.187 0.088 [ 2.992 7.523]\n",
"\t-5.47 1.163 0.086 [-7.699 -2.975]\n",
"\t\n",
"\t\n",
"\tPosterior quantiles:\n",
"\t\n",
"\t2.5 25 50 75 97.5\n",
"\t |---------------|===============|===============|---------------|\n",
"\t-1.611 -0.278 0.38 1.033 2.308\n",
"\t2.806 4.194 4.921 5.804 7.326\n",
"\t1.725 4.307 5.374 6.584 8.488\n",
"\t-7.774 -5.886 -5.104 -4.253 -2.817\n",
"\t-9.926 -7.507 -6.054 -4.984 -2.559\n",
"\t-18.693 -16.396 -14.435 -12.781 -9.379\n",
"\t2.521 5.466 6.854 8.135 11.346\n",
"\t-5.348 -4.725 -4.348 -4.007 -3.295\n",
"\t-2.252 -1.522 -1.125 -0.751 -0.052\n",
"\t2.763 4.226 5.077 5.989 7.354\n",
"\t-7.777 -6.235 -5.483 -4.715 -3.03\n",
"\t\n"
]
}
],
"source": [
"M_receptive_vocab.beta.summary()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Database queries"
]
},
{
"cell_type": "code",
"execution_count": 118,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"import sqlite3"
]
},
{
"cell_type": "code",
"execution_count": 119,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"db_files = !ls *sqlite"
]
},
{
"cell_type": "code",
"execution_count": 120,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"articulation = sqlite3.connect(\"articulation.sqlite\")"
]
},
{
"cell_type": "code",
"execution_count": 121,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"c_artic = articulation.cursor()"
]
},
{
"cell_type": "code",
"execution_count": 122,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"<sqlite3.Cursor at 0x11a8de6c0>"
]
},
"execution_count": 122,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"c_artic.execute('SELECT v1 FROM sigma_school WHERE trace==0')"
]
},
{
"cell_type": "code",
"execution_count": 123,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"sigma_school_artic = np.ravel(c_artic.fetchall())"
]
},
{
"cell_type": "code",
"execution_count": 124,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"5.7923510709000006"
]
},
"execution_count": 124,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sigma_school_artic.mean()"
]
},
{
"cell_type": "code",
"execution_count": 125,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"articulation.sqlite\n",
"5.739025\n",
"[ 3.090618 8.920649]\n",
"\n",
"explang.sqlite\n",
"4.945318\n",
"[ 3.390639 7.225201]\n",
"\n",
"expressive_vocab.sqlite\n",
"8.008659\n",
"[ 6.058134 10.607475]\n",
"\n",
"receptive_vocab.sqlite\n",
"6.532987\n",
"[ 4.738706 9.103903]\n",
"\n",
"reclang.sqlite\n",
"5.1217285\n",
"[ 3.251533 7.165217]\n"
]
}
],
"source": [
"for dbname in db_files:\n",
" db = sqlite3.connect(dbname)\n",
" cur = db.cursor()\n",
" cur.execute('SELECT v1 FROM sigma_school WHERE trace==0')\n",
" trace = np.ravel(cur.fetchall())\n",
" trace.sort()\n",
" print('\\n'+dbname)\n",
" print(np.median(trace))\n",
" print(trace[[int(len(trace)*0.025), int(len(trace)*0.975)]])"
]
}
],
"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": "3.4.2"
}
},
"nbformat": 4,
"nbformat_minor": 0
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment