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
<liclass="toctree-l3"><aclass="reference internal" href="#different-resolutions-between-model-and-observations">Different resolutions between model and observations</a></li>
@@ -161,11 +162,15 @@ <h2>Checking output<a class="headerlink" href="#checking-output" title="Link to
161
162
<p>If you are using a machine that allows both types of paths, it is recommended to list your <codeclass="docutils literal notranslate"><spanclass="pre">gpfs</span></code> paths first. This is because when reading data from the <codeclass="docutils literal notranslate"><spanclass="pre">esarchive</span></code>, a major limitation on the read time is the transfer speed between the 2 machines, reading directly from the <codeclass="docutils literal notranslate"><spanclass="pre">gpfs</span></code> directory circumvents this therefore.</p>
162
163
</section>
163
164
<sectionid="additional-considerations">
164
-
<h2>Additional Considerations<aclass="headerlink" href="#additional-considerations" title="Link to this heading"></a></h2>
165
+
<h2>Additional considerations<aclass="headerlink" href="#additional-considerations" title="Link to this heading"></a></h2>
165
166
<p>When checking if an experiment is stored in a location with the corresponding domain, resolution, and species, consider that the species might not always be listed under the same name.</p>
166
167
<p>The file <codeclass="docutils literal notranslate"><spanclass="pre">internal/mapping_species.yaml</span></code> contains a dictionary mapping original species names to their alternative names.</p>
167
168
<p>Note that the mapping species file is only used when the species name from the configuration file is not found in the expected location, meaning Povidentia first looks for the species written in the configuration file. If it is not found, it then searches for the corresponding mapped species in <codeclass="docutils literal notranslate"><spanclass="pre">mapping_species.yaml</span></code>.</p>
<h2>Different resolutions between model and observations<aclass="headerlink" href="#different-resolutions-between-model-and-observations" title="Link to this heading"></a></h2>
172
+
<p>When the resolutions are different the model data downsamples to the observations resolution. For instance, if we have 3 hourly model data and daily observations, the model data will be resampled to daily. For this to happen, we need to set the resolution in our configuration file as the obervations frequency.</p>
@@ -377,16 +379,16 @@ <h2>Create providentia-env_v2.4.0 in MN5<a class="headerlink" href="#create-prov
377
379
<sectionid="create-local-environment">
378
380
<h2>Create local environment<aclass="headerlink" href="#create-local-environment" title="Link to this heading"></a></h2>
379
381
<p>In order to test modules with pip, you need to create an environment. Once activated, you can start installing modules using either <codeclass="docutils literal notranslate"><spanclass="pre">pip</span></code> or <codeclass="docutils literal notranslate"><spanclass="pre">conda</span></code>, as in this example:</p>
<p>Here we create an environment called <codeclass="docutils literal notranslate"><spanclass="pre">providentia-env</span></code> with the Python version 3.11.5, and we install the latest version of Cartopy (with conda, with pip it gives problems), and the Python packages from <codeclass="docutils literal notranslate"><spanclass="pre">requirements.txt</span></code> using pip.</p>
391
+
<p>Here we create an environment called <codeclass="docutils literal notranslate"><spanclass="pre">providentia-env_v3.00</span></code> with the Python version 3.11.5, and we install the latest version of Cartopy (with conda, with pip it gives problems), and the Python packages from <codeclass="docutils literal notranslate"><spanclass="pre">requirements.txt</span></code> using pip.</p>
390
392
<p>pip install memray</p>
391
393
<p>memray run -o result.bin python your_script.py
392
394
memray flamegraph result.bin # Creates a flamegraph
0 commit comments