From 209657c50bc18257429c00e4d5f69afbce609fe5 Mon Sep 17 00:00:00 2001
From: Pieter Van Trappen <vtpieter@gmail.com>
Date: Thu, 23 Apr 2020 23:09:01 +0200
Subject: [PATCH] first changes to make it works, fails on missing
logbook_test.csv now
---
README2020.md | 15 ++++++++++-----
src/scripts/grid_search_gmm.py | 12 ++++++++----
2 files changed, 18 insertions(+), 9 deletions(-)
diff --git a/README2020.md b/README2020.md
index 72d3ddc8..b4c46339 100644
--- a/README2020.md
+++ b/README2020.md
@@ -1,7 +1,12 @@
-- pull ML_step
+# Pipeline instructions
-- grid_search_gmm.py en grid_search_iforest.py zouden beiden moeten werken en output produceren in de data-cern subfolder onder oftewel gmm_results of iforest_results
+* grid_search_gmm.py en grid_search_iforest.py zouden beiden moeten werken en output produceren in de data-cern subfolder onder oftewel gmm_results of iforest_results
+* grid_search_gmm.py gerbuikt parameters '-b 1' (zonder string haakjes). grid_search_iforest.py gebruikt er geen.
+* voor de webapp run wsgi_lbds.py in de webapp folder
-- grid_search_gmm.py gerbuikt parameters '-b 1' (zonder string haakjes). grid_search_iforest.py gebruikt er geen.
-
-- voor de webapp run wsgi_lbds.py in de webapp folder
+run as follows:
+```
+pipenv shell
+cd src/scripts
+./grid_search_gmm.py
+```
diff --git a/src/scripts/grid_search_gmm.py b/src/scripts/grid_search_gmm.py
index c87ad9aa..56250e15 100755
--- a/src/scripts/grid_search_gmm.py
+++ b/src/scripts/grid_search_gmm.py
@@ -2,6 +2,7 @@
import os
import sys
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
from pipeline.grid_search import grid_search
from preprocessing import builder
@@ -66,10 +67,13 @@ if __name__ == '__main__':
fno = "res"
df = pd.read_csv("../../data-cern/features.csv")
- state_mode = pd.read_json("../../data-cern/seg.json",
- lines=True)
- state_mode = state_mode.set_index('acqStamp').sort_index()
- state_mode.index = pd.to_datetime(state_mode.index).round('s')
+ try:
+ fn = "../../data-cern/seg.json"
+ state_mode = pd.read_json(fn, lines=True)
+ state_mode = state_mode.set_index('acqStamp').sort_index()
+ state_mode.index = pd.to_datetime(state_mode.index).round('s')
+ except:
+ print("ERROR - missing {}, continuing anyway".format(fn))
df = df.set_index('acqStamp').sort_index()
df.index = pd.to_datetime(df.index)
labels = pd.DataFrame()
--
GitLab