You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: Installing-PYME-with-py3-macos-ARM64.md
+32Lines changed: 32 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -1,5 +1,7 @@
1
1
## Instructions to install PYME natively on macos ARM64 (M1, M2 etc)
2
2
3
+
### Installing miniconda and xcode command line tools
4
+
3
5
Everything is generally done in a terminal window that runs natively.
4
6
5
7
We start with a native miniconda install:
@@ -13,6 +15,8 @@ That installed a small base install with Python 3.9.
13
15
14
16
Then I picked up the `xcode` command-line tools which is quite easy. There are a number of ways, a very easy one is just to type a command like `clang` at the terminal prompt and then you are immediately prompted/asked if you would like to install the xcode command line tools. See also https://www.freecodecamp.org/news/install-xcode-command-line-tools/.
15
17
18
+
### Installing PYME proper
19
+
16
20
I made an environment to build PYME in and based that on Python 3.8 as a well tested version according to David Baddeley:
17
21
18
22
```shell
@@ -50,3 +54,31 @@ dh5view -t
50
54
```
51
55
52
56
This seems to give a working install, still testing but most things with `dh5view` and `PYMEVis` seem fine.
57
+
58
+
### Getting jupyter notebook support going
59
+
60
+
This needed another quick conda install:
61
+
62
+
```shell
63
+
conda install -c conda-forge notebook
64
+
```
65
+
66
+
I then tried opening a new python3 notebook and got an error:
67
+
68
+
```
69
+
500 : Internal Server Error
70
+
```
71
+
72
+
A quick google led me to this [stackoverflow discussion](https://stackoverflow.com/questions/36851746/jupyter-notebook-500-internal-server-error) which suggested to run
73
+
74
+
```
75
+
conda install nbconvert==5.4.1
76
+
```
77
+
78
+
Now, I am not sure that the pinned version is still the best option (the previous conda command had installed version `nbconvert-5.6.1`), but I gave it a go anyway (leaving the option to tweak to a later version once it is clear in which version the issue occurs) and did
79
+
80
+
```shell
81
+
conda install -c conda-forge nbconvert==5.4.1
82
+
```
83
+
84
+
which indeed fixed the issue and now I can open notebooks just fine.
0 commit comments