Skip to content

Commit

Permalink
restaura dvc/config
Browse files Browse the repository at this point in the history
  • Loading branch information
@f3270
f3270 committed Oct 9, 2024
1 parent 59a6130 commit cb02c27
Show file tree
Hide file tree
Showing 2 changed files with 172 additions and 83 deletions.
4 changes: 4 additions & 0 deletions .dvc/config
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,4 @@
[core]
remote = s3-remote
['remote "s3-remote"']
url = s3://mlops-project-t6/mlops-data
251 changes: 168 additions & 83 deletions notebooks/Fase_1_Equipo6_v1.2_ftorres.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -951,43 +951,43 @@
{
"data": {
"text/plain": [
"Recipientgender 0\n",
"Stemcellsource 0\n",
"Donorage 0\n",
"Donorage35 0\n",
"IIIV 0\n",
"Gendermatch 0\n",
"DonorABO 0\n",
"RecipientABO 1\n",
"RecipientRh 2\n",
"ABOmatch 1\n",
"CMVstatus 16\n",
"DonorCMV 2\n",
"RecipientCMV 14\n",
"Disease 0\n",
"Riskgroup 0\n",
"Txpostrelapse 0\n",
"Diseasegroup 0\n",
"HLAmatch 0\n",
"HLAmismatch 0\n",
"Antigen 1\n",
"Allele 1\n",
"HLAgrI 0\n",
"Recipientage 0\n",
"Recipientage10 0\n",
"Recipientageint 0\n",
"Relapse 0\n",
"aGvHDIIIIV 0\n",
"extcGvHD 31\n",
"CD34kgx10d6 0\n",
"CD3dCD34 5\n",
"CD3dkgx10d8 5\n",
"Rbodymass 2\n",
"ANCrecovery 0\n",
"PLTrecovery 0\n",
"time_to_aGvHD_III_IV 0\n",
"survival_time 0\n",
"dtype: int64"
"Recipientgender 0.000000\n",
"Stemcellsource 0.000000\n",
"Donorage 0.000000\n",
"Donorage35 0.000000\n",
"IIIV 0.000000\n",
"Gendermatch 0.000000\n",
"DonorABO 0.000000\n",
"RecipientABO 0.534759\n",
"RecipientRh 1.069519\n",
"ABOmatch 0.534759\n",
"CMVstatus 8.556150\n",
"DonorCMV 1.069519\n",
"RecipientCMV 7.486631\n",
"Disease 0.000000\n",
"Riskgroup 0.000000\n",
"Txpostrelapse 0.000000\n",
"Diseasegroup 0.000000\n",
"HLAmatch 0.000000\n",
"HLAmismatch 0.000000\n",
"Antigen 0.534759\n",
"Allele 0.534759\n",
"HLAgrI 0.000000\n",
"Recipientage 0.000000\n",
"Recipientage10 0.000000\n",
"Recipientageint 0.000000\n",
"Relapse 0.000000\n",
"aGvHDIIIIV 0.000000\n",
"extcGvHD 16.577540\n",
"CD34kgx10d6 0.000000\n",
"CD3dCD34 2.673797\n",
"CD3dkgx10d8 2.673797\n",
"Rbodymass 1.069519\n",
"ANCrecovery 0.000000\n",
"PLTrecovery 0.000000\n",
"time_to_aGvHD_III_IV 0.000000\n",
"survival_time 0.000000\n",
"dtype: float64"
]
},
"execution_count": 13,
Expand All @@ -996,8 +996,8 @@
}
],
"source": [
"X.isna().sum()\n",
"# X.isna().sum() * 100 / len(X) # pct"
"# X.isna().sum()\n",
"X.isna().sum() * 100 / len(X) # pct"
]
},
{
Expand Down Expand Up @@ -1227,64 +1227,149 @@
},
{
"cell_type": "code",
"execution_count": 18,
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th>Stemcellsource</th>\n",
" <th>0</th>\n",
" <th>1</th>\n",
" </tr>\n",
" <tr>\n",
" <th>Recipientgender</th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>18</td>\n",
" <td>57</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>24</td>\n",
" <td>88</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
"Recipientgender 0\n",
"Stemcellsource 0\n",
"Donorage 0\n",
"Donorage35 0\n",
"IIIV 0\n",
"Gendermatch 0\n",
"DonorABO 0\n",
"RecipientABO 1\n",
"RecipientRh 2\n",
"ABOmatch 1\n",
"CMVstatus 16\n",
"DonorCMV 2\n",
"RecipientCMV 14\n",
"Disease 0\n",
"Riskgroup 0\n",
"Txpostrelapse 0\n",
"Diseasegroup 0\n",
"HLAmatch 0\n",
"HLAmismatch 0\n",
"Antigen 1\n",
"Allele 1\n",
"HLAgrI 0\n",
"Recipientage 0\n",
"Recipientage10 0\n",
"Recipientageint 0\n",
"Relapse 0\n",
"aGvHDIIIIV 0\n",
"extcGvHD 31\n",
"CD34kgx10d6 0\n",
"CD3dCD34 5\n",
"CD3dkgx10d8 5\n",
"Rbodymass 2\n",
"ANCrecovery 0\n",
"PLTrecovery 0\n",
"time_to_aGvHD_III_IV 0\n",
"survival_time 0\n",
"survival_status 0\n",
"dtype: int64"
"Stemcellsource 0 1\n",
"Recipientgender \n",
"0 18 57\n",
"1 24 88"
]
},
"execution_count": 18,
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.isna().sum()"
"pd.crosstab(df['Recipientgender'], df['Stemcellsource'])"
]
},
{
"cell_type": "code",
"execution_count": 17,
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th>Stemcellsource</th>\n",
" <th>0</th>\n",
" <th>1</th>\n",
" </tr>\n",
" <tr>\n",
" <th>Recipientgender</th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>18</td>\n",
" <td>57</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>24</td>\n",
" <td>88</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
"Stemcellsource 0 1\n",
"Recipientgender \n",
"0 18 57\n",
"1 24 88"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pd.crosstab(df[categoric_cols[0]],df[categoric_cols[1]])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# sns.pairplot(X[numeric_cols])\n",
"# plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -1301,11 +1386,11 @@
},
{
"cell_type": "code",
"execution_count": 19,
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df.to_csv('../data/raw/bone-marrow.csv', index=False, header=True)"
"# df.to_csv('../data/raw/bone-marrow.csv', index=False, header=True)"
]
},
{
Expand Down

0 comments on commit cb02c27

Please sign in to comment.